From Jayeeta.Basu@nyumc.org  Sat Sep  3 23:41:55 2016
Return-Path: <Jayeeta.Basu@nyumc.org>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u843ftZI013106
	for <shasha@mail.cims.nyu.edu>; Sat, 3 Sep 2016 23:41:55 -0400 (EDT)
Received: from mail7.nyumc.org (mail7.nyumc.org [216.165.127.151])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u843fnKi009114
	for <shasha@cims.nyu.edu>; Sat, 3 Sep 2016 23:41:49 -0400 (EDT)
X-IronPort-AV: E=Sophos;i="5.30,277,1470715200"; 
   d="scan'208,217";a="84151688"
Received: from msgwcdcpvmpr05.nyumc.org (HELO MSGWCDCPHT67.nyumc.org) ([10.185.58.191])
  by mail7.nyumc.org with SMTP; 03 Sep 2016 23:16:22 -0400
Received: from MSGWBDCPMBX32.nyumc.org ([169.254.9.9]) by
 MSGWCDCPHT67.nyumc.org ([169.254.36.34]) with mapi id 14.03.0174.001; Sat, 3
 Sep 2016 23:16:20 -0400
From: "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org>
To: Eftychios Pnevmatikakis <epnevmatikakis@simonsfoundation.org>
CC: "Zemla, Roland" <Roland.Zemla@med.nyu.edu>,
        "agiovannucci@simonsfoundation.org" <agiovannucci@simonsfoundation.org>,
        "dchklovskii@simonsfoundation.org" <dchklovskii@simonsfoundation.org>,
        "jkogan18@cgps.org" <jkogan18@cgps.org>,
        "shasha@cims.nyu.edu"
	<shasha@cims.nyu.edu>
Subject: Re: Somatic ROI and signal extraction 
Thread-Topic: Somatic ROI and signal extraction 
Thread-Index: AQHSBlq04GV/cjBMXkiCQxFqmScr0g==
Date: Sun, 4 Sep 2016 03:16:19 +0000
Message-ID: <fb766811-02ff-4b7a-82c6-ba9e6e03e087@MSGWCDCPHT67.nyumc.org>
References: <da940b3c-90ff-4807-a6ae-10ed679e6e09@MSGWBDCPVMHT62.nyumc.org>,<CAHcmRjCcU7wUdM0Weogz_8CP7h5kG8d_Z8W0jHp18nfLB_yBUA@mail.gmail.com>
In-Reply-To: <CAHcmRjCcU7wUdM0Weogz_8CP7h5kG8d_Z8W0jHp18nfLB_yBUA@mail.gmail.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Content-Type: multipart/alternative;
	boundary="_000_fb76681102ff4b7a82c6ba9e6e03e087MSGWCDCPHT67nyumcorg_"
MIME-Version: 1.0
X-CFilter-Loop: Reflected
X-Greylist: Delayed for 00:25:13 by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Sat, 03 Sep 2016 23:41:49 -0400 (EDT)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99
Status: R

--_000_fb76681102ff4b7a82c6ba9e6e03e087MSGWCDCPHT67nyumcorg_
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

Dear Eftichios and Giovanni
We are working with Dennis Sasha's group at Courant to look at some of the =
behavioral correlations and patterns of the calcium imaging data. They were=
 asking if we can provide them the Cartesian xyz coordinates for each ROI e=
xtracted from the t-series with your algorithm .

Do let us know if this information is readily available.

Best

Jayeeta

On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis <epnevmatikakis@simons=
foundation.org<mailto:epnevmatikakis@simonsfoundation.org>> wrote:

Hi Zemla,

Thanks a lot for sending these datasets. I applied our algorithm in a compl=
ete out of the box fashion for the 60Hz dataset and I get good denoising an=
d cell body capturing results, but it's harder to assess the performance fo=
r the dendrites. Some closer inspection will be needed there, although I st=
ill attach the script that I used. Unfortunately, I'll be out on medical le=
ave for the next 10-15 days starting tomorrow, so I'm not sure I can get to=
 that before that.

If I remember correctly, you also had issues with the motion correction fro=
m SIMA (or other sources) being very slow. We used calblitz (only on channe=
l 2) and it was able to correct for motion very fast and with good quality.=
 I'm attaching the code for this. Fast imaging helps here. I assume that fo=
r 30Hz things would also work, but it might be trickier for 15Hz due to int=
ra-frame motion. You can try it out and Andrea can help with any issues you=
're having. Once you get good motion correction up and running, I prefer yo=
u to just send me a single (motion corrected) multi-page .tiff file for fur=
ther analysis.

In general I think 6 minute experiments are probably too long. Shorter data=
sets could be sufficient although I see that the responses can be fairly sp=
arse. As you can see from my script I only used the first 4000 frames, alth=
ough I think that this is probably too little for 60Hz imaging rate and spa=
rse activation.

Hope that helps and we can talk more when I return. Andrea will be availabl=
e throughout this time for questions about motion correction and the algori=
thm.

Best,
Eftychios

PS: Apologies for the mix between matlab and python


On Tue, Jun 28, 2016 at 10:16 PM, Zemla, Roland <Roland.Zemla@med.nyu.edu<m=
ailto:Roland.Zemla@med.nyu.edu>> wrote:
Dear Andrea and Eftychios,

Thank you so much for taking the time to discuss the problems we are curren=
tly facing with imaging analysis at our last meeting. It was very helpful.

I am attaching data that I acquired today for the somatic/dendritic ROI ide=
ntification and signal extraction. I am currently working on acquiring axon=
al and bouton data as well.  I am curious to know how well the process can =
be automated with our existing data acquisition setup. Particularly, I woul=
d like to know what imaging rate you would suggest to be optimal for our ex=
periments and what the minimum duration of our imaging datasets would have =
to be given the sparsity of our signals (for use with your software).

Below are the links to 6+ minute continuous 2P resonance acquisitions at 15=
Hz (15.308), 30Hz (30.322), and 60Hz (59.955) with a 16x/0.8 NA objective a=
nd 2X optical zoom of the mouse running on the wheel and receiving water re=
ward every 40 'positions.' The signal is from GCamp6f-labeled pyramidal cel=
ls in CA1 of the hippocampus. The associated lick signal and run signal are=
 attached as a .csv file with each >100ms pulse >4.5V being a lick made by =
the mouse and each >100ms pulse >4.5V being a movement made by the mouse on=
 the wheel. The data is not motion corrected or post-processed in any way.

60Hz:
https://drive.google.com/file/d/0B412-7LDirTkZzN4dkJNRXBPbTg/view?usp=3Dsha=
ring<https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.co=
m_file_d_0B412-2D7LDirTkZzN4dkJNRXBPbTg_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=
=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GM=
cFoiF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DG7L=
iI5g9TNDJMy7WiHNbuk52FvFHvtFXT2fhl8ogLWk&e=3D>
30Hz:
https://drive.google.com/file/d/0B412-7LDirTkUDFMTkZjd1BWQ28/view?usp=3Dsha=
ring<https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.co=
m_file_d_0B412-2D7LDirTkUDFMTkZjd1BWQ28_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=
=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GM=
cFoiF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DeFj=
IZIdERvMgV4-lhmAW9doqbw8LpdCLLf9EXFexcxM&e=3D>
15Hz:
https://drive.google.com/file/d/0B412-7LDirTkTjBpZUV0ZVBGWnM/view?usp=3Dsha=
ring<https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.co=
m_file_d_0B412-2D7LDirTkTjBpZUV0ZVBGWnM_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=
=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GM=
cFoiF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DRGF=
wTrPBgSIWNj0-NCA0r2nqe5zejReyedj-TrQY2gc&e=3D>

Sincerely,

Roland Zemla
MD/PhD Candidate
Basu Laboratory
NYU Neuroscience Institute

------------------------------------------------------------
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D

<DemoMotionCorrection.py>
<demo_zemla.m>

------------------------------------------------------------
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D

--_000_fb76681102ff4b7a82c6ba9e6e03e087MSGWCDCPHT67nyumcorg_
Content-Type: text/html; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable

<html>
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii">
</head>
<body dir=3D"auto">
<div>Dear Eftichios and Giovanni</div>
<div id=3D"AppleMailSignature">We are working with Dennis Sasha's group at =
Courant to look at some of the behavioral correlations and patterns of the =
calcium imaging data. They were asking if we can provide them the Cartesian=
 xyz coordinates for each ROI extracted
 from the t-series with your algorithm .</div>
<div id=3D"AppleMailSignature"><br>
</div>
<div id=3D"AppleMailSignature">Do let us know if this information is readil=
y available.</div>
<div id=3D"AppleMailSignature"><br>
</div>
<div id=3D"AppleMailSignature">Best<br>
<br>
Jayeeta&nbsp;</div>
<div><br>
On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis &lt;<a href=3D"mailto:=
epnevmatikakis@simonsfoundation.org">epnevmatikakis@simonsfoundation.org</a=
>&gt; wrote:<br>
<br>
</div>
<blockquote type=3D"cite">
<div>
<div dir=3D"ltr">Hi Zemla,
<div><br>
</div>
<div>Thanks a lot for sending these datasets. I applied our algorithm in a =
complete out of the box fashion for the 60Hz dataset and I get good denoisi=
ng and cell body capturing results, but it's harder to assess the performan=
ce for the dendrites. Some closer
 inspection will be needed there, although I still attach the script that I=
 used. Unfortunately, I'll be out on medical leave for the next 10-15 days =
starting tomorrow, so I'm not sure I can get to that before that.</div>
<div><br>
</div>
<div>If I remember correctly, you also had issues with the motion correctio=
n from SIMA (or other sources) being very slow. We used calblitz (only on c=
hannel 2) and it was able to correct for motion very fast and with good qua=
lity. I'm attaching the code for
 this. Fast imaging helps here. I assume that for 30Hz things would also wo=
rk, but it might be trickier for 15Hz due to intra-frame motion. You can tr=
y it out and Andrea can help with any issues you're having. Once you get go=
od motion correction up and running,
 I prefer you to just send me a single (motion corrected) multi-page .tiff =
file for further analysis.&nbsp;</div>
<div><br>
</div>
<div>In general I think 6 minute experiments are probably too long. Shorter=
 datasets could be sufficient although I see that the responses can be fair=
ly sparse. As you can see from my script I only used the first 4000 frames,=
 although I think that this is probably
 too little for 60Hz imaging rate and sparse activation.&nbsp;</div>
<div><br>
</div>
<div>Hope that helps and we can talk more when I return. Andrea will be ava=
ilable throughout this time for questions about motion correction and the a=
lgorithm.&nbsp;</div>
<div><br>
</div>
<div>Best,&nbsp;</div>
<div>Eftychios</div>
<div><br>
</div>
<div>PS: Apologies for the mix between matlab and python</div>
<div><br>
</div>
</div>
<div class=3D"gmail_extra"><br>
<div class=3D"gmail_quote">On Tue, Jun 28, 2016 at 10:16 PM, Zemla, Roland =
<span dir=3D"ltr">
&lt;<a href=3D"mailto:Roland.Zemla@med.nyu.edu" target=3D"_blank">Roland.Ze=
mla@med.nyu.edu</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
<div>
<div style=3D"direction:ltr;font-family:Tahoma;color:#000000;font-size:10pt=
">
<div style=3D"direction:ltr"><span style=3D"font-size:10pt">Dear Andrea and=
 Eftychios,</span><br>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">Thank you=
 so much for taking the time to discuss the problems we are currently facin=
g with imaging analysis at our last meeting. It was very helpful.&nbsp;</di=
v>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">I am atta=
ching data that I acquired today for the somatic/dendritic ROI identificati=
on and signal extraction. I am currently working on acquiring axonal and bo=
uton data as well.&nbsp; I am curious to
 know how well the process can be automated with our existing data acquisit=
ion setup. Particularly, I would like to know what imaging rate you would s=
uggest to be optimal for our experiments and what the minimum duration of o=
ur imaging datasets would have to
 be given the sparsity of our signals (for use with your software).</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">Below are=
 the links to 6&#43; minute continuous 2P resonance acquisitions at 15Hz (1=
5.308), 30Hz (30.322), and 60Hz (59.955) with a 16x/0.8 NA objective and 2X=
 optical zoom of the mouse running on
 the wheel and receiving water reward every 40 'positions.' The signal is f=
rom GCamp6f-labeled pyramidal cells in CA1 of the hippocampus. The associat=
ed lick signal and run signal are attached as a .csv file with each &gt;100=
ms pulse &gt;4.5V being a lick made by
 the mouse and each&nbsp;<span style=3D"font-size:13.3333px">&gt;100ms puls=
e &gt;4.5V being a movement made by the mouse on the wheel. The data is not=
 motion corrected or post-processed in any way.</span></div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">60Hz:</di=
v>
<div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkZzN4dkJNRXBPbTg_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DG7LiI5g9TNDJMy7WiHNbuk52FvFHvtFXT2fhl8ogLWk&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/d/0B412-7LDirTkZzN4dkJ=
NRXBPbTg/view?usp=3Dsharing</a></div>
<div>30Hz:</div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkUDFMTkZjd1BWQ28_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DeFjIZIdERvMgV4-lhmAW9doqbw8LpdCLLf9EXFexcxM&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/d/0B412-7LDirTkUDFMTkZ=
jd1BWQ28/view?usp=3Dsharing</a></div>
<div>15Hz:</div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkTjBpZUV0ZVBGWnM_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DRGFwTrPBgSIWNj0-NCA0r2nqe5zejReyedj-TrQY2gc&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/d/0B412-7LDirTkTjBpZUV=
0ZVBGWnM/view?usp=3Dsharing</a></div>
<div><br>
</div>
<div>Sincerely,</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:13px">
<div>
<div>
<div>
<div>
<div>
<div style=3D"font-family:Tahoma;font-size:13px"><font size=3D"3" face=3D"V=
erdana"><br>
</font></div>
<div><font size=3D"2" face=3D"Arial"><font>Roland Zemla </font></font>
<div><font size=3D"2" face=3D"Arial">MD/PhD Candidate</font></div>
<div><font size=3D"2" face=3D"Arial">Basu Laboratory</font></div>
<div><font size=3D"2" face=3D"Arial">NYU Neuroscience Institute</font></div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<p><br>
------------------------------------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the re=
cipient should check this email and any attachments for the presence of vir=
uses. The organization accepts no
 liability for any damage caused by any virus transmitted by this email.<br>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D</p>
</div>
</blockquote>
</div>
<br>
</div>
</div>
</blockquote>
<blockquote type=3D"cite">
<div>&lt;DemoMotionCorrection.py&gt;</div>
</blockquote>
<blockquote type=3D"cite">
<div>&lt;demo_zemla.m&gt;</div>
</blockquote>
<p><br>
------------------------------------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.<br>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D</p></body>
</html>

--_000_fb76681102ff4b7a82c6ba9e6e03e087MSGWCDCPHT67nyumcorg_--

From agiovannucci@simonsfoundation.org  Sun Sep  4 06:08:58 2016
Return-Path: <agiovannucci@simonsfoundation.org>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u84A8ww6008885
	for <shasha@mail.cims.nyu.edu>; Sun, 4 Sep 2016 06:08:58 -0400 (EDT)
Received: from mail-it0-f44.google.com (mail-it0-f44.google.com [209.85.214.44])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u84A8nLX002078
	for <shasha@cims.nyu.edu>; Sun, 4 Sep 2016 06:08:49 -0400 (EDT)
Received: by mail-it0-f44.google.com with SMTP id c198so107486026ith.1
        for <shasha@cims.nyu.edu>; Sun, 04 Sep 2016 03:08:49 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=simonsfoundation.org; s=google;
        h=mime-version:in-reply-to:references:from:date:message-id:subject:to
         :cc;
        bh=4xN/eIKXPAg50AH3RvvIS6Hav+Br+cFNYCAfJE+LHtI=;
        b=HY5yvoRq9s+SkD8DV/v8s+lT7M+bmi0TgbJnPSla9rX3jx65yQo0DFTe6YBwZ4nn8r
         jqm5e5HWBhBRqWCKXf4yvcJcOqbU9Ni7dneb/Ghej1Rn3UrAbwXmyrMGqoOOLzlzEm6V
         6mZg6/kr8a/043SvVL5560zLCHRLr7W8MoucA=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=1e100.net; s=20130820;
        h=x-gm-message-state:mime-version:in-reply-to:references:from:date
         :message-id:subject:to:cc;
        bh=4xN/eIKXPAg50AH3RvvIS6Hav+Br+cFNYCAfJE+LHtI=;
        b=Q6MKbDh6msx1/qd4RNiATe0imiMWTL4UvnKOmJckBn1tBCzGHFf/4JxiqzFGuJqKrV
         4Znl1EWWSr3MYnP+d9xzLa6MURSud9Cgqu8fFz6euMC0Z8zHrr61OXtXYpYQYTPtxn2l
         a/cmsZhrQ/Xn5CD6xpipW8KIPWUPYWymTQoKTOAz+bWUP3EZmx1fNd0OpIqgjsZBi/jD
         O5pFse06sAWBA0m0jwQfkMltkRU3yKtrpfJg7jdCOwhEcEfJcv6p6QuSdue6DcDQVLMN
         cpI8Vksact3EfLSJSauQRQd6Tn+6d7GEVm0EEPvrpRP0Wb//uJubq7tu2QRKdX5cGSYt
         Uk6Q==
X-Gm-Message-State: AE9vXwNEwnv1vMA6ThxxRogcfRahRhD06eQ+2XKusM1hzkgJIXKFTg+PY3oUxYVYvK0NXzI4bntZzILZCL5n1mO1
X-Received: by 10.36.16.138 with SMTP id 132mr16505246ity.60.1472983724270;
 Sun, 04 Sep 2016 03:08:44 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.36.254.131 with HTTP; Sun, 4 Sep 2016 03:08:43 -0700 (PDT)
Received: by 10.36.254.131 with HTTP; Sun, 4 Sep 2016 03:08:43 -0700 (PDT)
In-Reply-To: <fb766811-02ff-4b7a-82c6-ba9e6e03e087@MSGWCDCPHT67.nyumc.org>
References: <da940b3c-90ff-4807-a6ae-10ed679e6e09@MSGWBDCPVMHT62.nyumc.org>
 <CAHcmRjCcU7wUdM0Weogz_8CP7h5kG8d_Z8W0jHp18nfLB_yBUA@mail.gmail.com> <fb766811-02ff-4b7a-82c6-ba9e6e03e087@MSGWCDCPHT67.nyumc.org>
From: Andrea Giovannucci <agiovannucci@simonsfoundation.org>
Date: Sun, 4 Sep 2016 06:08:43 -0400
Message-ID: <CA+mVXTaF=FGhrdsitw_GcYYPeX7bsfVi4oGzNiQh2AhOwBZXew@mail.gmail.com>
Subject: Re: Somatic ROI and signal extraction
To: "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org>
Cc: shasha@cims.nyu.edu, Dmitri Chklovskii <dchklovskii@simonsfoundation.org>,
        Eftychios Pnevmatikakis <epnevmatikakis@simonsfoundation.org>,
        jkogan18@cgps.org, "Zemla, Roland" <Roland.Zemla@med.nyu.edu>,
        Johannes Friedrich <jf2954@columbia.edu>
Content-Type: multipart/alternative; boundary=001a114382e8a54153053babbf6e
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Sun, 04 Sep 2016 06:08:49 -0400 (EDT)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99
Status: R

--001a114382e8a54153053babbf6e
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Dear Basu,

If you are using the Python version I would  extend the com function in the
utilities package. However in cc the person that implemented the 3D version
and might have some other advice. On Monday I will have a deeper look at it
as well.

Good luck,

Andrea

On Sep 4, 2016 05:16, "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org> wrote:

> Dear Eftichios and Giovanni
> We are working with Dennis Sasha's group at Courant to look at some of th=
e
> behavioral correlations and patterns of the calcium imaging data. They we=
re
> asking if we can provide them the Cartesian xyz coordinates for each ROI
> extracted from the t-series with your algorithm .
>
> Do let us know if this information is readily available.
>
> Best
>
> Jayeeta
>
> On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis <epnevmatikakis@
> simonsfoundation.org> wrote:
>
> Hi Zemla,
>
> Thanks a lot for sending these datasets. I applied our algorithm in a
> complete out of the box fashion for the 60Hz dataset and I get good
> denoising and cell body capturing results, but it's harder to assess the
> performance for the dendrites. Some closer inspection will be needed ther=
e,
> although I still attach the script that I used. Unfortunately, I'll be ou=
t
> on medical leave for the next 10-15 days starting tomorrow, so I'm not su=
re
> I can get to that before that.
>
> If I remember correctly, you also had issues with the motion correction
> from SIMA (or other sources) being very slow. We used calblitz (only on
> channel 2) and it was able to correct for motion very fast and with good
> quality. I'm attaching the code for this. Fast imaging helps here. I assu=
me
> that for 30Hz things would also work, but it might be trickier for 15Hz d=
ue
> to intra-frame motion. You can try it out and Andrea can help with any
> issues you're having. Once you get good motion correction up and running,=
 I
> prefer you to just send me a single (motion corrected) multi-page .tiff
> file for further analysis.
>
> In general I think 6 minute experiments are probably too long. Shorter
> datasets could be sufficient although I see that the responses can be
> fairly sparse. As you can see from my script I only used the first 4000
> frames, although I think that this is probably too little for 60Hz imagin=
g
> rate and sparse activation.
>
> Hope that helps and we can talk more when I return. Andrea will be
> available throughout this time for questions about motion correction and
> the algorithm.
>
> Best,
> Eftychios
>
> PS: Apologies for the mix between matlab and python
>
>
> On Tue, Jun 28, 2016 at 10:16 PM, Zemla, Roland <Roland.Zemla@med.nyu.edu=
>
> wrote:
>
>> Dear Andrea and Eftychios,
>>
>> Thank you so much for taking the time to discuss the problems we are
>> currently facing with imaging analysis at our last meeting. It was very
>> helpful.
>>
>> I am attaching data that I acquired today for the somatic/dendritic ROI
>> identification and signal extraction. I am currently working on acquirin=
g
>> axonal and bouton data as well.  I am curious to know how well the proce=
ss
>> can be automated with our existing data acquisition setup. Particularly,=
 I
>> would like to know what imaging rate you would suggest to be optimal for
>> our experiments and what the minimum duration of our imaging datasets wo=
uld
>> have to be given the sparsity of our signals (for use with your software=
).
>>
>> Below are the links to 6+ minute continuous 2P resonance acquisitions at
>> 15Hz (15.308), 30Hz (30.322), and 60Hz (59.955) with a 16x/0.8 NA object=
ive
>> and 2X optical zoom of the mouse running on the wheel and receiving wate=
r
>> reward every 40 'positions.' The signal is from GCamp6f-labeled pyramida=
l
>> cells in CA1 of the hippocampus. The associated lick signal and run sign=
al
>> are attached as a .csv file with each >100ms pulse >4.5V being a lick ma=
de
>> by the mouse and each >100ms pulse >4.5V being a movement made by the
>> mouse on the wheel. The data is not motion corrected or post-processed i=
n
>> any way.
>>
>> 60Hz:
>> https://drive.google.com/file/d/0B412-7LDirTkZzN4dkJNRXBPbTg/view?
>> usp=3Dsharing
>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.com=
_file_d_0B412-2D7LDirTkZzN4dkJNRXBPbTg_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=3D=
j5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GMcFo=
iF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DG7LiI5=
g9TNDJMy7WiHNbuk52FvFHvtFXT2fhl8ogLWk&e=3D>
>> 30Hz:
>> https://drive.google.com/file/d/0B412-7LDirTkUDFMTkZjd1BWQ28/view?
>> usp=3Dsharing
>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.com=
_file_d_0B412-2D7LDirTkUDFMTkZjd1BWQ28_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=3D=
j5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GMcFo=
iF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DeFjIZI=
dERvMgV4-lhmAW9doqbw8LpdCLLf9EXFexcxM&e=3D>
>> 15Hz:
>> https://drive.google.com/file/d/0B412-7LDirTkTjBpZUV0ZVBGWnM/view?
>> usp=3Dsharing
>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.com=
_file_d_0B412-2D7LDirTkTjBpZUV0ZVBGWnM_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=3D=
j5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GMcFo=
iF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DRGFwTr=
PBgSIWNj0-NCA0r2nqe5zejReyedj-TrQY2gc&e=3D>
>>
>> Sincerely,
>>
>> Roland Zemla
>> MD/PhD Candidate
>> Basu Laboratory
>> NYU Neuroscience Institute
>>
>>
>> ------------------------------------------------------------
>> This email message, including any attachments, is for the sole use of th=
e
>> intended recipient(s) and may contain information that is proprietary,
>> confidential, and exempt from disclosure under applicable law. Any
>> unauthorized review, use, disclosure, or distribution is prohibited. If =
you
>> have received this email in error please notify the sender by return ema=
il
>> and delete the original message. Please note, the recipient should check
>> this email and any attachments for the presence of viruses. The
>> organization accepts no liability for any damage caused by any virus
>> transmitted by this email.
>> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
>>
>
> <DemoMotionCorrection.py>
>
> <demo_zemla.m>
>
>
> ------------------------------------------------------------
> This email message, including any attachments, is for the sole use of the
> intended recipient(s) and may contain information that is proprietary,
> confidential, and exempt from disclosure under applicable law. Any
> unauthorized review, use, disclosure, or distribution is prohibited. If y=
ou
> have received this email in error please notify the sender by return emai=
l
> and delete the original message. Please note, the recipient should check
> this email and any attachments for the presence of viruses. The
> organization accepts no liability for any damage caused by any virus
> transmitted by this email.
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
>

--001a114382e8a54153053babbf6e
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<p dir=3D"ltr">Dear Basu,</p>
<p dir=3D"ltr">If you are using the Python version I would=C2=A0 extend the=
 com function in the utilities package. However in cc the person that imple=
mented the 3D version and might have some other advice. On Monday I will ha=
ve a deeper look at it as well. </p>
<p dir=3D"ltr">Good luck,</p>
<p dir=3D"ltr">Andrea</p>
<div class=3D"gmail_extra"><br><div class=3D"gmail_quote">On Sep 4, 2016 05=
:16, &quot;Basu, Jayeeta&quot; &lt;<a href=3D"mailto:Jayeeta.Basu@nyumc.org=
">Jayeeta.Basu@nyumc.org</a>&gt; wrote:<br type=3D"attribution"><blockquote=
 class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc soli=
d;padding-left:1ex">



<div dir=3D"auto">
<div>Dear Eftichios and Giovanni</div>
<div>We are working with Dennis Sasha&#39;s group at Courant to look at som=
e of the behavioral correlations and patterns of the calcium imaging data. =
They were asking if we can provide them the Cartesian xyz coordinates for e=
ach ROI extracted
 from the t-series with your algorithm .</div>
<div><br>
</div>
<div>Do let us know if this information is readily available.</div>
<div><br>
</div>
<div>Best<br>
<br>
Jayeeta=C2=A0</div>
<div><br>
On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis &lt;<a href=3D"mailto:=
epnevmatikakis@simonsfoundation.org" target=3D"_blank">epnevmatikakis@<wbr>=
simonsfoundation.org</a>&gt; wrote:<br>
<br>
</div>
<blockquote type=3D"cite">
<div>
<div dir=3D"ltr">Hi Zemla,
<div><br>
</div>
<div>Thanks a lot for sending these datasets. I applied our algorithm in a =
complete out of the box fashion for the 60Hz dataset and I get good denoisi=
ng and cell body capturing results, but it&#39;s harder to assess the perfo=
rmance for the dendrites. Some closer
 inspection will be needed there, although I still attach the script that I=
 used. Unfortunately, I&#39;ll be out on medical leave for the next 10-15 d=
ays starting tomorrow, so I&#39;m not sure I can get to that before that.</=
div>
<div><br>
</div>
<div>If I remember correctly, you also had issues with the motion correctio=
n from SIMA (or other sources) being very slow. We used calblitz (only on c=
hannel 2) and it was able to correct for motion very fast and with good qua=
lity. I&#39;m attaching the code for
 this. Fast imaging helps here. I assume that for 30Hz things would also wo=
rk, but it might be trickier for 15Hz due to intra-frame motion. You can tr=
y it out and Andrea can help with any issues you&#39;re having. Once you ge=
t good motion correction up and running,
 I prefer you to just send me a single (motion corrected) multi-page .tiff =
file for further analysis.=C2=A0</div>
<div><br>
</div>
<div>In general I think 6 minute experiments are probably too long. Shorter=
 datasets could be sufficient although I see that the responses can be fair=
ly sparse. As you can see from my script I only used the first 4000 frames,=
 although I think that this is probably
 too little for 60Hz imaging rate and sparse activation.=C2=A0</div>
<div><br>
</div>
<div>Hope that helps and we can talk more when I return. Andrea will be ava=
ilable throughout this time for questions about motion correction and the a=
lgorithm.=C2=A0</div>
<div><br>
</div>
<div>Best,=C2=A0</div>
<div>Eftychios</div>
<div><br>
</div>
<div>PS: Apologies for the mix between matlab and python</div>
<div><br>
</div>
</div>
<div class=3D"gmail_extra"><br>
<div class=3D"gmail_quote">On Tue, Jun 28, 2016 at 10:16 PM, Zemla, Roland =
<span dir=3D"ltr">
&lt;<a href=3D"mailto:Roland.Zemla@med.nyu.edu" target=3D"_blank">Roland.Ze=
mla@med.nyu.edu</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
<div>
<div style=3D"direction:ltr;font-family:Tahoma;color:#000000;font-size:10pt=
">
<div style=3D"direction:ltr"><span style=3D"font-size:10pt">Dear Andrea and=
 Eftychios,</span><br>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">Thank you=
 so much for taking the time to discuss the problems we are currently facin=
g with imaging analysis at our last meeting. It was very helpful.=C2=A0</di=
v>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">I am atta=
ching data that I acquired today for the somatic/dendritic ROI identificati=
on and signal extraction. I am currently working on acquiring axonal and bo=
uton data as well.=C2=A0 I am curious to
 know how well the process can be automated with our existing data acquisit=
ion setup. Particularly, I would like to know what imaging rate you would s=
uggest to be optimal for our experiments and what the minimum duration of o=
ur imaging datasets would have to
 be given the sparsity of our signals (for use with your software).</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">Below are=
 the links to 6+ minute continuous 2P resonance acquisitions at 15Hz (15.30=
8), 30Hz (30.322), and 60Hz (59.955) with a 16x/0.8 NA objective and 2X opt=
ical zoom of the mouse running on
 the wheel and receiving water reward every 40 &#39;positions.&#39; The sig=
nal is from GCamp6f-labeled pyramidal cells in CA1 of the hippocampus. The =
associated lick signal and run signal are attached as a .csv file with each=
 &gt;100ms pulse &gt;4.5V being a lick made by
 the mouse and each=C2=A0<span style=3D"font-size:13.3333px">&gt;100ms puls=
e &gt;4.5V being a movement made by the mouse on the wheel. The data is not=
 motion corrected or post-processed in any way.</span></div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">60Hz:</di=
v>
<div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkZzN4dkJNRXBPbTg_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DG7LiI5g9TNDJMy7WiHNbuk52FvFHvtFXT2fhl8ogLWk&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/<wbr>d/0B412-<wbr>7LDi=
rTkZzN4dkJNRXBPbTg/view?<wbr>usp=3Dsharing</a></div>
<div>30Hz:</div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkUDFMTkZjd1BWQ28_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DeFjIZIdERvMgV4-lhmAW9doqbw8LpdCLLf9EXFexcxM&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/<wbr>d/0B412-<wbr>7LDi=
rTkUDFMTkZjd1BWQ28/view?<wbr>usp=3Dsharing</a></div>
<div>15Hz:</div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkTjBpZUV0ZVBGWnM_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DRGFwTrPBgSIWNj0-NCA0r2nqe5zejReyedj-TrQY2gc&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/<wbr>d/0B412-<wbr>7LDi=
rTkTjBpZUV0ZVBGWnM/view?<wbr>usp=3Dsharing</a></div>
<div><br>
</div>
<div>Sincerely,</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:13px">
<div>
<div>
<div>
<div>
<div>
<div style=3D"font-family:Tahoma;font-size:13px"><font size=3D"3" face=3D"V=
erdana"><br>
</font></div>
<div><font size=3D"2" face=3D"Arial"><font>Roland Zemla </font></font>
<div><font size=3D"2" face=3D"Arial">MD/PhD Candidate</font></div>
<div><font size=3D"2" face=3D"Arial">Basu Laboratory</font></div>
<div><font size=3D"2" face=3D"Arial">NYU Neuroscience Institute</font></div=
>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<p><br>
------------------------------<wbr>------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the re=
cipient should check this email and any attachments for the presence of vir=
uses. The organization accepts no
 liability for any damage caused by any virus transmitted by this email.<br=
>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr>=3D=3D=3D</p>
</div>
</blockquote>
</div>
<br>
</div>
</div>
</blockquote>
<blockquote type=3D"cite">
<div>&lt;DemoMotionCorrection.py&gt;</div>
</blockquote>
<blockquote type=3D"cite">
<div>&lt;demo_zemla.m&gt;</div>
</blockquote>
<p><br>
------------------------------<wbr>------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.<br>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr>=3D=3D=3D</p></div>

</blockquote></div></div>

--001a114382e8a54153053babbf6e--

From epnevmatikakis@simonsfoundation.org  Mon Sep  5 12:33:55 2016
Return-Path: <epnevmatikakis@simonsfoundation.org>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u85GXtZu019194
	for <shasha@mail.cims.nyu.edu>; Mon, 5 Sep 2016 12:33:55 -0400 (EDT)
Received: from mail-wm0-f48.google.com (mail-wm0-f48.google.com [74.125.82.48])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u85GXlqo010939
	for <shasha@cims.nyu.edu>; Mon, 5 Sep 2016 12:33:47 -0400 (EDT)
Received: by mail-wm0-f48.google.com with SMTP id b187so27066416wme.1
        for <shasha@cims.nyu.edu>; Mon, 05 Sep 2016 09:33:47 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=simonsfoundation.org; s=google;
        h=mime-version:in-reply-to:references:from:date:message-id:subject:to
         :cc;
        bh=9vqVP9fynvFS4A13iGDRRFV56qntblGCQHKOc4sqYqQ=;
        b=fRR99S9dhmYcIu+GHZCD4XRMDeDvwA599Env3NjFpO+i3me11KItadKioBAFs65uC7
         98dkWCicPMwikRRj8W+kQt5RfG6KxFw8YdMbbWxwwLoNmltMyLDSs2MvVidsbRpJpdwM
         uZ+xiggZ2HCaQRfJ41LeOXN+cFtOqnAhQ0p5c=
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=1e100.net; s=20130820;
        h=x-gm-message-state:mime-version:in-reply-to:references:from:date
         :message-id:subject:to:cc;
        bh=9vqVP9fynvFS4A13iGDRRFV56qntblGCQHKOc4sqYqQ=;
        b=Bgs2EAiCX+DEDT6F+XsmqJcdSdQS3a5I01EaqK12agex/pysNulBfadsC58BmuAZNd
         5/zLYmLbMY+u+sSd+7kaBItHloxdCpV6MYpiX/n6GmLvpmt71Jwt1Pa/XLWlPCuU/CBF
         9U/ZIP7sWbs0Rdqa5seEfOl9qa/OKPQy/xYQ440eEDVxa2CaoD3/kAV87ZmZFCChGolr
         0HQaJGzVcCDvqLXcoVaWcPQD30FJqt/ae1rlWLNgI32rM+b/KmZxDe4iWMdLmz8mlV64
         F/AdTEBGAAl/MWZftCBKPaKZNwdfMk7xS0zyQCCzZdF+/Jlc00DkfE6C6jMy2x6sH7Zt
         jbyA==
X-Gm-Message-State: AE9vXwPcQC+q/b/i5mMkesNIlcS2ksWolieufWDC0wOk0VosRpTwhbj8ZPn2Pu9CH7917BHrkBx/s2qh93ylR8eZ
X-Received: by 10.194.27.202 with SMTP id v10mr30532479wjg.140.1473093221782;
 Mon, 05 Sep 2016 09:33:41 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.28.214.71 with HTTP; Mon, 5 Sep 2016 09:33:41 -0700 (PDT)
In-Reply-To: <0901cecf-f57a-41e4-8325-928305aa14cd@MSGWBDCPVMHT61.nyumc.org>
References: <da940b3c-90ff-4807-a6ae-10ed679e6e09@MSGWBDCPVMHT62.nyumc.org>
 <CAHcmRjCcU7wUdM0Weogz_8CP7h5kG8d_Z8W0jHp18nfLB_yBUA@mail.gmail.com>
 <fb766811-02ff-4b7a-82c6-ba9e6e03e087@MSGWCDCPHT67.nyumc.org>
 <CA+mVXTaF=FGhrdsitw_GcYYPeX7bsfVi4oGzNiQh2AhOwBZXew@mail.gmail.com> <0901cecf-f57a-41e4-8325-928305aa14cd@MSGWBDCPVMHT61.nyumc.org>
From: Eftychios Pnevmatikakis <epnevmatikakis@simonsfoundation.org>
Date: Mon, 5 Sep 2016 19:33:41 +0300
Message-ID: <CAHcmRjDMU-8YssnD_1dKBOPXjhM19TN3h1KFFy-L=CuhwCJhtQ@mail.gmail.com>
Subject: Re: Somatic ROI and signal extraction
To: "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org>
Cc: Andrea Giovannucci <agiovannucci@simonsfoundation.org>,
        "shasha@cims.nyu.edu" <shasha@cims.nyu.edu>,
        Dmitri Chklovskii <dchklovskii@simonsfoundation.org>,
        "jkogan18@cgps.org" <jkogan18@cgps.org>,
        "Zemla, Roland" <Roland.Zemla@med.nyu.edu>,
        Johannes Friedrich <jf2954@columbia.edu>
Content-Type: multipart/alternative; boundary=047d7b670a7b35092f053bc53e3c
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Mon, 05 Sep 2016 12:33:48 -0400 (EDT)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99
Status: R

--047d7b670a7b35092f053bc53e3c
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Dear Basu,

I'm not sure whether you ended up using matlab or python for the analysis.

In matlab you can use the com
<https://github.com/epnev/ca_source_extraction/blob/master/utilities/com.m>
function to extract the center of mass (centroid for each component).
You can also use the function plot_components
<https://github.com/epnev/ca_source_extraction/blob/master/utilities/plot_c=
ontours.m>
to
extract contour plots for each component.

A python we have similar functions:
There is a com function for the center of mass and a plot_contours for the
contour plots.
Both of them are in the utilities
<https://github.com/agiovann/Constrained_NMF/blob/master/ca_source_extracti=
on/utilities.py>
function.

Take a look at the demo files to see how they are used.

Thanks,
Eftychios



On Sun, Sep 4, 2016 at 2:14 PM, Basu, Jayeeta <Jayeeta.Basu@nyumc.org>
wrote:

> Dear Andrea
> Thanks for following up! Look forward to your inputs on Monday
> Best
>
> Jayeeta
>
> On Sep 4, 2016, at 6:08 AM, Andrea Giovannucci <agiovannucci@
> simonsfoundation.org> wrote:
>
> Dear Basu,
>
> If you are using the Python version I would  extend the com function in
> the utilities package. However in cc the person that implemented the 3D
> version and might have some other advice. On Monday I will have a deeper
> look at it as well.
>
> Good luck,
>
> Andrea
>
> On Sep 4, 2016 05:16, "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org> wrote:
>
>> Dear Eftichios and Giovanni
>> We are working with Dennis Sasha's group at Courant to look at some of
>> the behavioral correlations and patterns of the calcium imaging data. Th=
ey
>> were asking if we can provide them the Cartesian xyz coordinates for eac=
h
>> ROI extracted from the t-series with your algorithm .
>>
>> Do let us know if this information is readily available.
>>
>> Best
>>
>> Jayeeta
>>
>> On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis <
>> epnevmatikakis@simonsfoundation.org> wrote:
>>
>> Hi Zemla,
>>
>> Thanks a lot for sending these datasets. I applied our algorithm in a
>> complete out of the box fashion for the 60Hz dataset and I get good
>> denoising and cell body capturing results, but it's harder to assess the
>> performance for the dendrites. Some closer inspection will be needed the=
re,
>> although I still attach the script that I used. Unfortunately, I'll be o=
ut
>> on medical leave for the next 10-15 days starting tomorrow, so I'm not s=
ure
>> I can get to that before that.
>>
>> If I remember correctly, you also had issues with the motion correction
>> from SIMA (or other sources) being very slow. We used calblitz (only on
>> channel 2) and it was able to correct for motion very fast and with good
>> quality. I'm attaching the code for this. Fast imaging helps here. I ass=
ume
>> that for 30Hz things would also work, but it might be trickier for 15Hz =
due
>> to intra-frame motion. You can try it out and Andrea can help with any
>> issues you're having. Once you get good motion correction up and running=
, I
>> prefer you to just send me a single (motion corrected) multi-page .tiff
>> file for further analysis.
>>
>> In general I think 6 minute experiments are probably too long. Shorter
>> datasets could be sufficient although I see that the responses can be
>> fairly sparse. As you can see from my script I only used the first 4000
>> frames, although I think that this is probably too little for 60Hz imagi=
ng
>> rate and sparse activation.
>>
>> Hope that helps and we can talk more when I return. Andrea will be
>> available throughout this time for questions about motion correction and
>> the algorithm.
>>
>> Best,
>> Eftychios
>>
>> PS: Apologies for the mix between matlab and python
>>
>>
>> On Tue, Jun 28, 2016 at 10:16 PM, Zemla, Roland <Roland.Zemla@med.nyu.ed=
u
>> > wrote:
>>
>>> Dear Andrea and Eftychios,
>>>
>>> Thank you so much for taking the time to discuss the problems we are
>>> currently facing with imaging analysis at our last meeting. It was very
>>> helpful.
>>>
>>> I am attaching data that I acquired today for the somatic/dendritic ROI
>>> identification and signal extraction. I am currently working on acquiri=
ng
>>> axonal and bouton data as well.  I am curious to know how well the proc=
ess
>>> can be automated with our existing data acquisition setup. Particularly=
, I
>>> would like to know what imaging rate you would suggest to be optimal fo=
r
>>> our experiments and what the minimum duration of our imaging datasets w=
ould
>>> have to be given the sparsity of our signals (for use with your softwar=
e).
>>>
>>> Below are the links to 6+ minute continuous 2P resonance acquisitions a=
t
>>> 15Hz (15.308), 30Hz (30.322), and 60Hz (59.955) with a 16x/0.8 NA objec=
tive
>>> and 2X optical zoom of the mouse running on the wheel and receiving wat=
er
>>> reward every 40 'positions.' The signal is from GCamp6f-labeled pyramid=
al
>>> cells in CA1 of the hippocampus. The associated lick signal and run sig=
nal
>>> are attached as a .csv file with each >100ms pulse >4.5V being a lick m=
ade
>>> by the mouse and each >100ms pulse >4.5V being a movement made by the
>>> mouse on the wheel. The data is not motion corrected or post-processed =
in
>>> any way.
>>>
>>> 60Hz:
>>> https://drive.google.com/file/d/0B412-7LDirTkZzN4dkJNRXBPbTg
>>> /view?usp=3Dsharing
>>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.co=
m_file_d_0B412-2D7LDirTkZzN4dkJNRXBPbTg_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=
=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GM=
cFoiF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DG7L=
iI5g9TNDJMy7WiHNbuk52FvFHvtFXT2fhl8ogLWk&e=3D>
>>> 30Hz:
>>> https://drive.google.com/file/d/0B412-7LDirTkUDFMTkZjd1BWQ28
>>> /view?usp=3Dsharing
>>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.co=
m_file_d_0B412-2D7LDirTkUDFMTkZjd1BWQ28_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=
=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GM=
cFoiF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DeFj=
IZIdERvMgV4-lhmAW9doqbw8LpdCLLf9EXFexcxM&e=3D>
>>> 15Hz:
>>> https://drive.google.com/file/d/0B412-7LDirTkTjBpZUV0ZVBGWnM
>>> /view?usp=3Dsharing
>>> <https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.google.co=
m_file_d_0B412-2D7LDirTkTjBpZUV0ZVBGWnM_view-3Fusp-3Dsharing&d=3DCwMFaQ&c=
=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&r=3Dk3ybVka5SJOl76RLO8VNy1GM=
cFoiF-463RN3SBJmj5o&m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_gPZrcdAMKRGKK0&s=3DRGF=
wTrPBgSIWNj0-NCA0r2nqe5zejReyedj-TrQY2gc&e=3D>
>>>
>>> Sincerely,
>>>
>>> Roland Zemla
>>> MD/PhD Candidate
>>> Basu Laboratory
>>> NYU Neuroscience Institute
>>>
>>>
>>> ------------------------------------------------------------
>>> This email message, including any attachments, is for the sole use of
>>> the intended recipient(s) and may contain information that is proprieta=
ry,
>>> confidential, and exempt from disclosure under applicable law. Any
>>> unauthorized review, use, disclosure, or distribution is prohibited. If=
 you
>>> have received this email in error please notify the sender by return em=
ail
>>> and delete the original message. Please note, the recipient should chec=
k
>>> this email and any attachments for the presence of viruses. The
>>> organization accepts no liability for any damage caused by any virus
>>> transmitted by this email.
>>> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
>>>
>>
>> <DemoMotionCorrection.py>
>>
>> <demo_zemla.m>
>>
>>
>> ------------------------------------------------------------
>> This email message, including any attachments, is for the sole use of th=
e
>> intended recipient(s) and may contain information that is proprietary,
>> confidential, and exempt from disclosure under applicable law. Any
>> unauthorized review, use, disclosure, or distribution is prohibited. If =
you
>> have received this email in error please notify the sender by return ema=
il
>> and delete the original message. Please note, the recipient should check
>> this email and any attachments for the presence of viruses. The
>> organization accepts no liability for any damage caused by any virus
>> transmitted by this email.
>> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
>>
>
> ------------------------------------------------------------
> This email message, including any attachments, is for the sole use of the
> intended recipient(s) and may contain information that is proprietary,
> confidential, and exempt from disclosure under applicable law. Any
> unauthorized review, use, disclosure, or distribution is prohibited. If y=
ou
> have received this email in error please notify the sender by return emai=
l
> and delete the original message. Please note, the recipient should check
> this email and any attachments for the presence of viruses. The
> organization accepts no liability for any damage caused by any virus
> transmitted by this email.
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
>

--047d7b670a7b35092f053bc53e3c
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Dear Basu,<div><br></div><div>I&#39;m not sure whether you=
 ended up using matlab or python for the analysis.=C2=A0</div><div><br></di=
v><div>In matlab you can use the <a href=3D"https://github.com/epnev/ca_sou=
rce_extraction/blob/master/utilities/com.m">com</a> function to extract the=
 center of mass (centroid for each component).=C2=A0</div><div>You can also=
 use the function <a href=3D"https://github.com/epnev/ca_source_extraction/=
blob/master/utilities/plot_contours.m">plot_components</a>=C2=A0to extract =
contour plots for each component.=C2=A0</div><div><br></div><div>A python w=
e have similar functions:</div><div>There is a com function for the center =
of mass and a plot_contours for the contour plots.</div><div>Both of them a=
re in the <a href=3D"https://github.com/agiovann/Constrained_NMF/blob/maste=
r/ca_source_extraction/utilities.py">utilities</a> function.=C2=A0</div><di=
v><br></div><div>Take a look at the demo files to see how they are used.</d=
iv><div><br></div><div>Thanks,</div><div>Eftychios=C2=A0</div><div><br></di=
v><div>=C2=A0</div></div><div class=3D"gmail_extra"><br><div class=3D"gmail=
_quote">On Sun, Sep 4, 2016 at 2:14 PM, Basu, Jayeeta <span dir=3D"ltr">&lt=
;<a href=3D"mailto:Jayeeta.Basu@nyumc.org" target=3D"_blank">Jayeeta.Basu@n=
yumc.org</a>&gt;</span> wrote:<br><blockquote class=3D"gmail_quote" style=
=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">



<div dir=3D"auto">
<div>Dear Andrea</div>
<div>Thanks for following up! Look forward to your inputs on Monday</div>
<div>Best<span class=3D"HOEnZb"><font color=3D"#888888"><br>
<br>
Jayeeta=C2=A0</font></span></div><div><div class=3D"h5">
<div><br>
On Sep 4, 2016, at 6:08 AM, Andrea Giovannucci &lt;<a href=3D"mailto:agiova=
nnucci@simonsfoundation.org" target=3D"_blank">agiovannucci@<wbr>simonsfoun=
dation.org</a>&gt; wrote:<br>
<br>
</div>
<blockquote type=3D"cite">
<div>
<p dir=3D"ltr">Dear Basu,</p>
<p dir=3D"ltr">If you are using the Python version I would=C2=A0 extend the=
 com function in the utilities package. However in cc the person that imple=
mented the 3D version and might have some other advice. On Monday I will ha=
ve a deeper look at it as well.
</p>
<p dir=3D"ltr">Good luck,</p>
<p dir=3D"ltr">Andrea</p>
<div class=3D"gmail_extra"><br>
<div class=3D"gmail_quote">On Sep 4, 2016 05:16, &quot;Basu, Jayeeta&quot; =
&lt;<a href=3D"mailto:Jayeeta.Basu@nyumc.org" target=3D"_blank">Jayeeta.Bas=
u@nyumc.org</a>&gt; wrote:<br type=3D"attribution">
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
<div dir=3D"auto">
<div>Dear Eftichios and Giovanni</div>
<div>We are working with Dennis Sasha&#39;s group at Courant to look at som=
e of the behavioral correlations and patterns of the calcium imaging data. =
They were asking if we can provide them the Cartesian xyz coordinates for e=
ach ROI extracted from the t-series
 with your algorithm .</div>
<div><br>
</div>
<div>Do let us know if this information is readily available.</div>
<div><br>
</div>
<div>Best<br>
<br>
Jayeeta=C2=A0</div>
<div><br>
On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis &lt;<a href=3D"mailto:=
epnevmatikakis@simonsfoundation.org" target=3D"_blank">epnevmatikakis@simon=
sfoundati<wbr>on.org</a>&gt; wrote:<br>
<br>
</div>
<blockquote type=3D"cite">
<div>
<div dir=3D"ltr">Hi Zemla,
<div><br>
</div>
<div>Thanks a lot for sending these datasets. I applied our algorithm in a =
complete out of the box fashion for the 60Hz dataset and I get good denoisi=
ng and cell body capturing results, but it&#39;s harder to assess the perfo=
rmance for the dendrites. Some closer
 inspection will be needed there, although I still attach the script that I=
 used. Unfortunately, I&#39;ll be out on medical leave for the next 10-15 d=
ays starting tomorrow, so I&#39;m not sure I can get to that before that.</=
div>
<div><br>
</div>
<div>If I remember correctly, you also had issues with the motion correctio=
n from SIMA (or other sources) being very slow. We used calblitz (only on c=
hannel 2) and it was able to correct for motion very fast and with good qua=
lity. I&#39;m attaching the code for
 this. Fast imaging helps here. I assume that for 30Hz things would also wo=
rk, but it might be trickier for 15Hz due to intra-frame motion. You can tr=
y it out and Andrea can help with any issues you&#39;re having. Once you ge=
t good motion correction up and running,
 I prefer you to just send me a single (motion corrected) multi-page .tiff =
file for further analysis.=C2=A0</div>
<div><br>
</div>
<div>In general I think 6 minute experiments are probably too long. Shorter=
 datasets could be sufficient although I see that the responses can be fair=
ly sparse. As you can see from my script I only used the first 4000 frames,=
 although I think that this is probably
 too little for 60Hz imaging rate and sparse activation.=C2=A0</div>
<div><br>
</div>
<div>Hope that helps and we can talk more when I return. Andrea will be ava=
ilable throughout this time for questions about motion correction and the a=
lgorithm.=C2=A0</div>
<div><br>
</div>
<div>Best,=C2=A0</div>
<div>Eftychios</div>
<div><br>
</div>
<div>PS: Apologies for the mix between matlab and python</div>
<div><br>
</div>
</div>
<div class=3D"gmail_extra"><br>
<div class=3D"gmail_quote">On Tue, Jun 28, 2016 at 10:16 PM, Zemla, Roland =
<span dir=3D"ltr">
&lt;<a href=3D"mailto:Roland.Zemla@med.nyu.edu" target=3D"_blank">Roland.Ze=
mla@med.nyu.edu</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
<div>
<div style=3D"direction:ltr;font-family:Tahoma;color:#000000;font-size:10pt=
">
<div style=3D"direction:ltr"><span style=3D"font-size:10pt">Dear Andrea and=
 Eftychios,</span><br>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">Thank you=
 so much for taking the time to discuss the problems we are currently facin=
g with imaging analysis at our last meeting. It was very helpful.=C2=A0</di=
v>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">I am atta=
ching data that I acquired today for the somatic/dendritic ROI identificati=
on and signal extraction. I am currently working on acquiring axonal and bo=
uton data as well.=C2=A0 I am curious to
 know how well the process can be automated with our existing data acquisit=
ion setup. Particularly, I would like to know what imaging rate you would s=
uggest to be optimal for our experiments and what the minimum duration of o=
ur imaging datasets would have to
 be given the sparsity of our signals (for use with your software).</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">Below are=
 the links to 6+ minute continuous 2P resonance acquisitions at 15Hz (15.30=
8), 30Hz (30.322), and 60Hz (59.955) with a 16x/0.8 NA objective and 2X opt=
ical zoom of the mouse running on
 the wheel and receiving water reward every 40 &#39;positions.&#39; The sig=
nal is from GCamp6f-labeled pyramidal cells in CA1 of the hippocampus. The =
associated lick signal and run signal are attached as a .csv file with each=
 &gt;100ms pulse &gt;4.5V being a lick made by
 the mouse and each=C2=A0<span style=3D"font-size:13.3333px">&gt;100ms puls=
e &gt;4.5V being a movement made by the mouse on the wheel. The data is not=
 motion corrected or post-processed in any way.</span></div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt"><br>
</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:10pt">60Hz:</di=
v>
<div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkZzN4dkJNRXBPbTg_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DG7LiI5g9TNDJMy7WiHNbuk52FvFHvtFXT2fhl8ogLWk&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/<wbr>d/0B412-7LDirTkZz=
N4dkJNRXBPbTg<wbr>/view?usp=3Dsharing</a></div>
<div>30Hz:</div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkUDFMTkZjd1BWQ28_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DeFjIZIdERvMgV4-lhmAW9doqbw8LpdCLLf9EXFexcxM&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/<wbr>d/0B412-7LDirTkUD=
FMTkZjd1BWQ28<wbr>/view?usp=3Dsharing</a></div>
<div>15Hz:</div>
<div><a href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__driv=
e.google.com_file_d_0B412-2D7LDirTkTjBpZUV0ZVBGWnM_view-3Fusp-3Dsharing&amp=
;d=3DCwMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3y=
bVka5SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9d=
dj_gPZrcdAMKRGKK0&amp;s=3DRGFwTrPBgSIWNj0-NCA0r2nqe5zejReyedj-TrQY2gc&amp;e=
=3D" target=3D"_blank">https://drive.google.com/file/<wbr>d/0B412-7LDirTkTj=
BpZUV0ZVBGWnM<wbr>/view?usp=3Dsharing</a></div>
<div><br>
</div>
<div>Sincerely,</div>
<div style=3D"color:rgb(0,0,0);font-family:Tahoma;font-size:13px">
<div>
<div>
<div>
<div>
<div>
<div style=3D"font-family:Tahoma;font-size:13px"><font size=3D"3" face=3D"V=
erdana"><br>
</font></div>
<div><font size=3D"2" face=3D"Arial"><font>Roland Zemla </font></font>
<div><font size=3D"2" face=3D"Arial">MD/PhD Candidate</font></div>
<div><font size=3D"2" face=3D"Arial">Basu Laboratory</font></div>
<div><font size=3D"2" face=3D"Arial">NYU Neuroscience Institute</font></div=
>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<p><br>
------------------------------<wbr>------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the re=
cipient should check this email and any attachments for the presence of vir=
uses. The organization accepts no
 liability for any damage caused by any virus transmitted by this email.<br=
>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr>=3D=3D=3D</p>
</div>
</blockquote>
</div>
<br>
</div>
</div>
</blockquote>
<blockquote type=3D"cite">
<div>&lt;DemoMotionCorrection.py&gt;</div>
</blockquote>
<blockquote type=3D"cite">
<div>&lt;demo_zemla.m&gt;</div>
</blockquote>
<p><br>
------------------------------<wbr>------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the re=
cipient should check this email and any attachments for the presence of vir=
uses. The organization accepts no
 liability for any damage caused by any virus transmitted by this email.<br=
>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr>=3D=3D=3D</p>
</div>
</blockquote>
</div>
</div>
</div>
</blockquote>
<p><br>
------------------------------<wbr>------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.<br>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr>=3D=3D=3D</p></div></div></div>

</blockquote></div><br></div>

--047d7b670a7b35092f053bc53e3c--

From jf2954@columbia.edu  Mon Sep  5 13:21:13 2016
Return-Path: <jf2954@columbia.edu>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u85HLDZY028328
	for <shasha@mail.cims.nyu.edu>; Mon, 5 Sep 2016 13:21:13 -0400 (EDT)
Received: from buckwheat.cc.columbia.edu (buckwheat.cc.columbia.edu [128.59.72.251])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u85HL5Jc015899
	for <shasha@cims.nyu.edu>; Mon, 5 Sep 2016 13:21:06 -0400 (EDT)
Received: from hazelnut (hazelnut.cc.columbia.edu [128.59.213.250])
	by buckwheat.cc.columbia.edu (8.13.8/8.13.8) with ESMTP id u85HG61Z019982
	for <shasha@cims.nyu.edu>; Mon, 5 Sep 2016 13:21:05 -0400
Received: from hazelnut (localhost.localdomain [127.0.0.1])
	by hazelnut (Postfix) with ESMTP id C5E9D6D
	for <shasha@cims.nyu.edu>; Mon,  5 Sep 2016 13:21:05 -0400 (EDT)
Received: from sendprodmail01.cc.columbia.edu (sendprodmail01.cc.columbia.edu [128.59.72.13])
	by hazelnut (Postfix) with ESMTP id AD6706D
	for <shasha@cims.nyu.edu>; Mon,  5 Sep 2016 13:21:05 -0400 (EDT)
Received: from mail-ua0-f198.google.com (mail-ua0-f198.google.com [209.85.217.198])
	by sendprodmail01.cc.columbia.edu (8.14.4/8.14.4) with ESMTP id u85HL5q8061805
	(version=TLSv1/SSLv3 cipher=AES128-GCM-SHA256 bits=128 verify=NOT)
	for <shasha@cims.nyu.edu>; Mon, 5 Sep 2016 13:21:05 -0400
Received: by mail-ua0-f198.google.com with SMTP id i32so399788163uai.0
        for <shasha@cims.nyu.edu>; Mon, 05 Sep 2016 10:21:05 -0700 (PDT)
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=1e100.net; s=20130820;
        h=x-gm-message-state:mime-version:subject:from:in-reply-to:date:cc
         :message-id:references:to;
        bh=lmLdQKkwSL41jE8Qb5P+SKqOlM+w36DDI2Oon2aTzJM=;
        b=STtB9PC0FzddKCJyN3uV+WXjx0eOOTN33cnXUHA4+u3Iaaq0vrHGXdZI0eB6+22n/U
         RtJwaALdYGYdo6tGrbEldGBbXVZowqNg4A6HAdR/fzL0AbZo5b3he05BCDjv1wUNDckx
         JZiNRW3XH3naeklaR7TV3ZD5XWkdo2BrHusAM8CirqKfM5BISyuT5ZHTMwam4DD9zhym
         Nzky144Kor3QxsQIA1nfn1zm2YJ8VjfnOC/5pELrRMbkEu7Na6KbuD6FtXWElXFhqaQy
         f0I07YQ9eg6l5fEBfmOo0WTI8JzNS9Vs+ArsK3/fvS9lJ8vxflYQk3w/YmaRBcsDi8l2
         ivQA==
X-Gm-Message-State: AE9vXwMhZ+5Lw4taZ5uXttAJ+P4meDQ+61zU6z+epjflBByRyYndYEGZlHt6IjytNn4Ji65nUBmfyY81YM/CHmaQ21mU+sofMdsoqoSa79ghJB6wvP3aA0PDpZ0KMqij2imG346QYBj9kdM=
X-Received: by 10.55.88.199 with SMTP id m190mr40342838qkb.78.1473096064810;
        Mon, 05 Sep 2016 10:21:04 -0700 (PDT)
X-Received: by 10.55.88.199 with SMTP id m190mr40342814qkb.78.1473096064589;
        Mon, 05 Sep 2016 10:21:04 -0700 (PDT)
Received: from [192.168.1.101] (dyn-128-59-109-45.dyn.columbia.edu. [128.59.109.45])
        by smtp.gmail.com with ESMTPSA id v43sm13577520qtv.15.2016.09.05.10.21.00
        (version=TLS1 cipher=ECDHE-RSA-AES128-SHA bits=128/128);
        Mon, 05 Sep 2016 10:21:02 -0700 (PDT)
Content-Type: multipart/alternative; boundary="Apple-Mail=_EC756431-9D33-46BD-9B34-F6C9F7D8193D"
Mime-Version: 1.0 (Mac OS X Mail 9.3 \(3124\))
Subject: Re: Somatic ROI and signal extraction
From: Johannes Friedrich <jf2954@columbia.edu>
In-Reply-To: <CAHcmRjDMU-8YssnD_1dKBOPXjhM19TN3h1KFFy-L=CuhwCJhtQ@mail.gmail.com>
Date: Mon, 5 Sep 2016 13:20:59 -0400
Cc: "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org>,
        Andrea Giovannucci <agiovannucci@simonsfoundation.org>,
        "shasha@cims.nyu.edu" <shasha@cims.nyu.edu>,
        Dmitri Chklovskii <dchklovskii@simonsfoundation.org>,
        "jkogan18@cgps.org" <jkogan18@cgps.org>,
        "Zemla, Roland" <Roland.Zemla@med.nyu.edu>
Message-Id: <8ADDF2B6-79FC-49B0-A759-B965744ECAAD@columbia.edu>
References: <da940b3c-90ff-4807-a6ae-10ed679e6e09@MSGWBDCPVMHT62.nyumc.org> <CAHcmRjCcU7wUdM0Weogz_8CP7h5kG8d_Z8W0jHp18nfLB_yBUA@mail.gmail.com> <fb766811-02ff-4b7a-82c6-ba9e6e03e087@MSGWCDCPHT67.nyumc.org> <CA+mVXTaF=FGhrdsitw_GcYYPeX7bsfVi4oGzNiQh2AhOwBZXew@mail.gmail.com> <0901cecf-f57a-41e4-8325-928305aa14cd@MSGWBDCPVMHT61.nyumc.org> <CAHcmRjDMU-8YssnD_1dKBOPXjhM19TN3h1KFFy-L=CuhwCJhtQ@mail.gmail.com>
To: Eftychios Pnevmatikakis <epnevmatikakis@simonsfoundation.org>
X-Mailer: Apple Mail (2.3124)
X-No-Spam-Score: Local
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99
X-Scanned-By: MIMEDefang 2.78 on 128.59.72.13
X-Greylist: Sender DNS name whitelisted, not delayed by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Mon, 05 Sep 2016 13:21:06 -0400 (EDT)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
Status: R


--Apple-Mail=_EC756431-9D33-46BD-9B34-F6C9F7D8193D
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=utf-8

Hi Basu,
in python you can use scipy.ndimage.measurements.center_of_mass, which =
works for 2D as well as 3D data (but only for dense arrays).=20
I didn=E2=80=99t bother to adapt the com function in utilities to the 3D =
case, as it=E2=80=99s already in ndimage. Was there a reason to write =
your own version, Eftychios?
Best,
Johannes

> On Sep 5, 2016, at 12:33 PM, Eftychios Pnevmatikakis =
<epnevmatikakis@simonsfoundation.org> wrote:
>=20
> Dear Basu,
>=20
> I'm not sure whether you ended up using matlab or python for the =
analysis.=20
>=20
> In matlab you can use the com =
<https://github.com/epnev/ca_source_extraction/blob/master/utilities/com.m=
> function to extract the center of mass (centroid for each component).=20=

> You can also use the function plot_components =
<https://github.com/epnev/ca_source_extraction/blob/master/utilities/plot_=
contours.m> to extract contour plots for each component.=20
>=20
> A python we have similar functions:
> There is a com function for the center of mass and a plot_contours for =
the contour plots.
> Both of them are in the utilities =
<https://github.com/agiovann/Constrained_NMF/blob/master/ca_source_extract=
ion/utilities.py> function.=20
>=20
> Take a look at the demo files to see how they are used.
>=20
> Thanks,
> Eftychios=20
>=20
> =20
>=20
> On Sun, Sep 4, 2016 at 2:14 PM, Basu, Jayeeta <Jayeeta.Basu@nyumc.org =
<mailto:Jayeeta.Basu@nyumc.org>> wrote:
> Dear Andrea
> Thanks for following up! Look forward to your inputs on Monday
> Best
>=20
> Jayeeta=20
>=20
> On Sep 4, 2016, at 6:08 AM, Andrea Giovannucci =
<agiovannucci@simonsfoundation.org =
<mailto:agiovannucci@simonsfoundation.org>> wrote:
>=20
>> Dear Basu,
>>=20
>> If you are using the Python version I would  extend the com function =
in the utilities package. However in cc the person that implemented the =
3D version and might have some other advice. On Monday I will have a =
deeper look at it as well.
>>=20
>> Good luck,
>>=20
>> Andrea
>>=20
>>=20
>> On Sep 4, 2016 05:16, "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org =
<mailto:Jayeeta.Basu@nyumc.org>> wrote:
>> Dear Eftichios and Giovanni
>> We are working with Dennis Sasha's group at Courant to look at some =
of the behavioral correlations and patterns of the calcium imaging data. =
They were asking if we can provide them the Cartesian xyz coordinates =
for each ROI extracted from the t-series with your algorithm .
>>=20
>> Do let us know if this information is readily available.
>>=20
>> Best
>>=20
>> Jayeeta=20
>>=20
>> On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis =
<epnevmatikakis@simonsfoundation.org =
<mailto:epnevmatikakis@simonsfoundation.org>> wrote:
>>=20
>>> Hi Zemla,
>>>=20
>>> Thanks a lot for sending these datasets. I applied our algorithm in =
a complete out of the box fashion for the 60Hz dataset and I get good =
denoising and cell body capturing results, but it's harder to assess the =
performance for the dendrites. Some closer inspection will be needed =
there, although I still attach the script that I used. Unfortunately, =
I'll be out on medical leave for the next 10-15 days starting tomorrow, =
so I'm not sure I can get to that before that.
>>>=20
>>> If I remember correctly, you also had issues with the motion =
correction from SIMA (or other sources) being very slow. We used =
calblitz (only on channel 2) and it was able to correct for motion very =
fast and with good quality. I'm attaching the code for this. Fast =
imaging helps here. I assume that for 30Hz things would also work, but =
it might be trickier for 15Hz due to intra-frame motion. You can try it =
out and Andrea can help with any issues you're having. Once you get good =
motion correction up and running, I prefer you to just send me a single =
(motion corrected) multi-page .tiff file for further analysis.=20
>>>=20
>>> In general I think 6 minute experiments are probably too long. =
Shorter datasets could be sufficient although I see that the responses =
can be fairly sparse. As you can see from my script I only used the =
first 4000 frames, although I think that this is probably too little for =
60Hz imaging rate and sparse activation.=20
>>>=20
>>> Hope that helps and we can talk more when I return. Andrea will be =
available throughout this time for questions about motion correction and =
the algorithm.=20
>>>=20
>>> Best,=20
>>> Eftychios
>>>=20
>>> PS: Apologies for the mix between matlab and python
>>>=20
>>>=20
>>> On Tue, Jun 28, 2016 at 10:16 PM, Zemla, Roland =
<Roland.Zemla@med.nyu.edu <mailto:Roland.Zemla@med.nyu.edu>> wrote:
>>> Dear Andrea and Eftychios,
>>>=20
>>> Thank you so much for taking the time to discuss the problems we are =
currently facing with imaging analysis at our last meeting. It was very =
helpful.=20
>>>=20
>>> I am attaching data that I acquired today for the somatic/dendritic =
ROI identification and signal extraction. I am currently working on =
acquiring axonal and bouton data as well.  I am curious to know how well =
the process can be automated with our existing data acquisition setup. =
Particularly, I would like to know what imaging rate you would suggest =
to be optimal for our experiments and what the minimum duration of our =
imaging datasets would have to be given the sparsity of our signals (for =
use with your software).
>>>=20
>>> Below are the links to 6+ minute continuous 2P resonance =
acquisitions at 15Hz (15.308), 30Hz (30.322), and 60Hz (59.955) with a =
16x/0.8 NA objective and 2X optical zoom of the mouse running on the =
wheel and receiving water reward every 40 'positions.' The signal is =
from GCamp6f-labeled pyramidal cells in CA1 of the hippocampus. The =
associated lick signal and run signal are attached as a .csv file with =
each >100ms pulse >4.5V being a lick made by the mouse and each >100ms =
pulse >4.5V being a movement made by the mouse on the wheel. The data is =
not motion corrected or post-processed in any way.
>>>=20
>>> 60Hz:
>>> =
https://drive.google.com/file/d/0B412-7LDirTkZzN4dkJNRXBPbTg/view?usp=3Dsh=
aring
>>> 30Hz:
>>> =
https://drive.google.com/file/d/0B412-7LDirTkUDFMTkZjd1BWQ28/view?usp=3Dsh=
aring
>>> 15Hz:
>>> =
https://drive.google.com/file/d/0B412-7LDirTkTjBpZUV0ZVBGWnM/view?usp=3Dsh=
aring
>>>=20
>>> Sincerely,
>>>=20
>>> Roland Zemla
>>> MD/PhD Candidate
>>> Basu Laboratory
>>> NYU Neuroscience Institute
>>>=20
>>> ------------------------------------------------------------
>>> This email message, including any attachments, is for the sole use =
of the intended recipient(s) and may contain information that is =
proprietary, confidential, and exempt from disclosure under applicable =
law. Any unauthorized review, use, disclosure, or distribution is =
prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the =
recipient should check this email and any attachments for the presence =
of viruses. The organization accepts no liability for any damage caused =
by any virus transmitted by this email.
>>> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
>>>=20
>>>=20
>>> <DemoMotionCorrection.py>
>>> <demo_zemla.m>
>>=20
>> ------------------------------------------------------------
>> This email message, including any attachments, is for the sole use of =
the intended recipient(s) and may contain information that is =
proprietary, confidential, and exempt from disclosure under applicable =
law. Any unauthorized review, use, disclosure, or distribution is =
prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the =
recipient should check this email and any attachments for the presence =
of viruses. The organization accepts no liability for any damage caused =
by any virus transmitted by this email.
>> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
>>=20
>=20
> ------------------------------------------------------------
> This email message, including any attachments, is for the sole use of =
the intended recipient(s) and may contain information that is =
proprietary, confidential, and exempt from disclosure under applicable =
law. Any unauthorized review, use, disclosure, or distribution is =
prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the =
recipient should check this email and any attachments for the presence =
of viruses. The organization accepts no liability for any damage caused =
by any virus transmitted by this email.
> =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D
>=20
>=20


--Apple-Mail=_EC756431-9D33-46BD-9B34-F6C9F7D8193D
Content-Transfer-Encoding: quoted-printable
Content-Type: text/html;
	charset=utf-8

<html><head><meta http-equiv=3D"Content-Type" content=3D"text/html =
charset=3Dutf-8"></head><body style=3D"word-wrap: break-word; =
-webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" =
class=3D""><div class=3D"">Hi Basu,</div><div class=3D"">in python you =
can use&nbsp;scipy.ndimage.measurements.center_of_mass, which works for =
2D as well as 3D data (but only for dense arrays).&nbsp;</div><div =
class=3D"">I didn=E2=80=99t bother to adapt the com function in =
utilities to the 3D case, as it=E2=80=99s already in ndimage. Was there =
a reason to write your own version, Eftychios?</div><div =
class=3D"">Best,</div><div class=3D"">Johannes</div><br =
class=3D""><div><blockquote type=3D"cite" class=3D""><div class=3D"">On =
Sep 5, 2016, at 12:33 PM, Eftychios Pnevmatikakis &lt;<a =
href=3D"mailto:epnevmatikakis@simonsfoundation.org" =
class=3D"">epnevmatikakis@simonsfoundation.org</a>&gt; wrote:</div><br =
class=3D"Apple-interchange-newline"><div class=3D""><div dir=3D"ltr" =
class=3D"">Dear Basu,<div class=3D""><br class=3D""></div><div =
class=3D"">I'm not sure whether you ended up using matlab or python for =
the analysis.&nbsp;</div><div class=3D""><br class=3D""></div><div =
class=3D"">In matlab you can use the <a =
href=3D"https://github.com/epnev/ca_source_extraction/blob/master/utilitie=
s/com.m" class=3D"">com</a> function to extract the center of mass =
(centroid for each component).&nbsp;</div><div class=3D"">You can also =
use the function <a =
href=3D"https://github.com/epnev/ca_source_extraction/blob/master/utilitie=
s/plot_contours.m" class=3D"">plot_components</a>&nbsp;to extract =
contour plots for each component.&nbsp;</div><div class=3D""><br =
class=3D""></div><div class=3D"">A python we have similar =
functions:</div><div class=3D"">There is a com function for the center =
of mass and a plot_contours for the contour plots.</div><div =
class=3D"">Both of them are in the <a =
href=3D"https://github.com/agiovann/Constrained_NMF/blob/master/ca_source_=
extraction/utilities.py" class=3D"">utilities</a> =
function.&nbsp;</div><div class=3D""><br class=3D""></div><div =
class=3D"">Take a look at the demo files to see how they are =
used.</div><div class=3D""><br class=3D""></div><div =
class=3D"">Thanks,</div><div class=3D"">Eftychios&nbsp;</div><div =
class=3D""><br class=3D""></div><div class=3D"">&nbsp;</div></div><div =
class=3D"gmail_extra"><br class=3D""><div class=3D"gmail_quote">On Sun, =
Sep 4, 2016 at 2:14 PM, Basu, Jayeeta <span dir=3D"ltr" class=3D"">&lt;<a =
href=3D"mailto:Jayeeta.Basu@nyumc.org" target=3D"_blank" =
class=3D"">Jayeeta.Basu@nyumc.org</a>&gt;</span> wrote:<br =
class=3D""><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 =
.8ex;border-left:1px #ccc solid;padding-left:1ex">



<div dir=3D"auto" class=3D"">
<div class=3D"">Dear Andrea</div>
<div class=3D"">Thanks for following up! Look forward to your inputs on =
Monday</div>
<div class=3D"">Best<span class=3D"HOEnZb"><font color=3D"#888888" =
class=3D""><br class=3D"">
<br class=3D"">
Jayeeta&nbsp;</font></span></div><div class=3D""><div class=3D"h5">
<div class=3D""><br class=3D"">
On Sep 4, 2016, at 6:08 AM, Andrea Giovannucci &lt;<a =
href=3D"mailto:agiovannucci@simonsfoundation.org" target=3D"_blank" =
class=3D"">agiovannucci@<wbr class=3D"">simonsfoundation.org</a>&gt; =
wrote:<br class=3D"">
<br class=3D"">
</div>
<blockquote type=3D"cite" class=3D"">
<div class=3D""><p dir=3D"ltr" class=3D"">Dear Basu,</p><p dir=3D"ltr" =
class=3D"">If you are using the Python version I would&nbsp; extend the =
com function in the utilities package. However in cc the person that =
implemented the 3D version and might have some other advice. On Monday I =
will have a deeper look at it as well.
</p><p dir=3D"ltr" class=3D"">Good luck,</p><p dir=3D"ltr" =
class=3D"">Andrea</p>
<div class=3D"gmail_extra"><br class=3D"">
<div class=3D"gmail_quote">On Sep 4, 2016 05:16, "Basu, Jayeeta" &lt;<a =
href=3D"mailto:Jayeeta.Basu@nyumc.org" target=3D"_blank" =
class=3D"">Jayeeta.Basu@nyumc.org</a>&gt; wrote:<br type=3D"attribution" =
class=3D"">
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 =
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div dir=3D"auto" class=3D"">
<div class=3D"">Dear Eftichios and Giovanni</div>
<div class=3D"">We are working with Dennis Sasha's group at Courant to =
look at some of the behavioral correlations and patterns of the calcium =
imaging data. They were asking if we can provide them the Cartesian xyz =
coordinates for each ROI extracted from the t-series
 with your algorithm .</div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">Do let us know if this information is readily =
available.</div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">Best<br class=3D"">
<br class=3D"">
Jayeeta&nbsp;</div>
<div class=3D""><br class=3D"">
On Jun 30, 2016, at 5:32 PM, Eftychios Pnevmatikakis &lt;<a =
href=3D"mailto:epnevmatikakis@simonsfoundation.org" target=3D"_blank" =
class=3D"">epnevmatikakis@simonsfoundati<wbr class=3D"">on.org</a>&gt; =
wrote:<br class=3D"">
<br class=3D"">
</div>
<blockquote type=3D"cite" class=3D"">
<div class=3D"">
<div dir=3D"ltr" class=3D"">Hi Zemla,
<div class=3D""><br class=3D"">
</div>
<div class=3D"">Thanks a lot for sending these datasets. I applied our =
algorithm in a complete out of the box fashion for the 60Hz dataset and =
I get good denoising and cell body capturing results, but it's harder to =
assess the performance for the dendrites. Some closer
 inspection will be needed there, although I still attach the script =
that I used. Unfortunately, I'll be out on medical leave for the next =
10-15 days starting tomorrow, so I'm not sure I can get to that before =
that.</div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">If I remember correctly, you also had issues with the =
motion correction from SIMA (or other sources) being very slow. We used =
calblitz (only on channel 2) and it was able to correct for motion very =
fast and with good quality. I'm attaching the code for
 this. Fast imaging helps here. I assume that for 30Hz things would also =
work, but it might be trickier for 15Hz due to intra-frame motion. You =
can try it out and Andrea can help with any issues you're having. Once =
you get good motion correction up and running,
 I prefer you to just send me a single (motion corrected) multi-page =
.tiff file for further analysis.&nbsp;</div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">In general I think 6 minute experiments are probably too =
long. Shorter datasets could be sufficient although I see that the =
responses can be fairly sparse. As you can see from my script I only =
used the first 4000 frames, although I think that this is probably
 too little for 60Hz imaging rate and sparse activation.&nbsp;</div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">Hope that helps and we can talk more when I return. =
Andrea will be available throughout this time for questions about motion =
correction and the algorithm.&nbsp;</div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">Best,&nbsp;</div>
<div class=3D"">Eftychios</div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">PS: Apologies for the mix between matlab and =
python</div>
<div class=3D""><br class=3D"">
</div>
</div>
<div class=3D"gmail_extra"><br class=3D"">
<div class=3D"gmail_quote">On Tue, Jun 28, 2016 at 10:16 PM, Zemla, =
Roland <span dir=3D"ltr" class=3D"">
&lt;<a href=3D"mailto:Roland.Zemla@med.nyu.edu" target=3D"_blank" =
class=3D"">Roland.Zemla@med.nyu.edu</a>&gt;</span> wrote:<br class=3D"">
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 =
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class=3D"">
<div style=3D"direction: ltr; font-family: Tahoma; font-size: 10pt;" =
class=3D"">
<div style=3D"direction:ltr" class=3D""><span style=3D"font-size:10pt" =
class=3D"">Dear Andrea and Eftychios,</span><br class=3D"">
<div style=3D"font-family: Tahoma; font-size: 10pt;" class=3D""><br =
class=3D"">
</div>
<div style=3D"font-family: Tahoma; font-size: 10pt;" class=3D"">Thank =
you so much for taking the time to discuss the problems we are currently =
facing with imaging analysis at our last meeting. It was very =
helpful.&nbsp;</div>
<div style=3D"font-family: Tahoma; font-size: 10pt;" class=3D""><br =
class=3D"">
</div>
<div style=3D"font-family: Tahoma; font-size: 10pt;" class=3D"">I am =
attaching data that I acquired today for the somatic/dendritic ROI =
identification and signal extraction. I am currently working on =
acquiring axonal and bouton data as well.&nbsp; I am curious to
 know how well the process can be automated with our existing data =
acquisition setup. Particularly, I would like to know what imaging rate =
you would suggest to be optimal for our experiments and what the minimum =
duration of our imaging datasets would have to
 be given the sparsity of our signals (for use with your =
software).</div>
<div style=3D"font-family: Tahoma; font-size: 10pt;" class=3D""><br =
class=3D"">
</div>
<div style=3D"font-family: Tahoma; font-size: 10pt;" class=3D"">Below =
are the links to 6+ minute continuous 2P resonance acquisitions at 15Hz =
(15.308), 30Hz (30.322), and 60Hz (59.955) with a 16x/0.8 NA objective =
and 2X optical zoom of the mouse running on
 the wheel and receiving water reward every 40 'positions.' The signal =
is from GCamp6f-labeled pyramidal cells in CA1 of the hippocampus. The =
associated lick signal and run signal are attached as a .csv file with =
each &gt;100ms pulse &gt;4.5V being a lick made by
 the mouse and each&nbsp;<span style=3D"font-size:13.3333px" =
class=3D"">&gt;100ms pulse &gt;4.5V being a movement made by the mouse =
on the wheel. The data is not motion corrected or post-processed in any =
way.</span></div>
<div style=3D"font-family: Tahoma; font-size: 10pt;" class=3D""><br =
class=3D"">
</div>
<div style=3D"font-family: Tahoma; font-size: 10pt;" =
class=3D"">60Hz:</div>
<div class=3D"">
<div class=3D""><!-- <a =
href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.googl=
e.com_file_d_0B412-2D7LDirTkZzN4dkJNRXBPbTg_view-3Fusp-3Dsharing&amp;d=3DC=
wMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3ybVka5=
SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_g=
PZrcdAMKRGKK0&amp;s=3DG7LiI5g9TNDJMy7WiHNbuk52FvFHvtFXT2fhl8ogLWk&amp;e=3D=
" target=3D"_blank"> --><a href=3D"https://drive.google.com/file/" =
class=3D"">https://drive.google.com/file/</a><wbr =
class=3D"">d/0B412-7LDirTkZzN4dkJNRXBPbTg<wbr =
class=3D"">/view?usp=3Dsharing<!-- </a> --></div>
<div class=3D"">30Hz:</div>
<div class=3D""><!-- <a =
href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.googl=
e.com_file_d_0B412-2D7LDirTkUDFMTkZjd1BWQ28_view-3Fusp-3Dsharing&amp;d=3DC=
wMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3ybVka5=
SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_g=
PZrcdAMKRGKK0&amp;s=3DeFjIZIdERvMgV4-lhmAW9doqbw8LpdCLLf9EXFexcxM&amp;e=3D=
" target=3D"_blank"> --><a href=3D"https://drive.google.com/file/" =
class=3D"">https://drive.google.com/file/</a><wbr =
class=3D"">d/0B412-7LDirTkUDFMTkZjd1BWQ28<wbr =
class=3D"">/view?usp=3Dsharing<!-- </a> --></div>
<div class=3D"">15Hz:</div>
<div class=3D""><!-- <a =
href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttps-3A__drive.googl=
e.com_file_d_0B412-2D7LDirTkTjBpZUV0ZVBGWnM_view-3Fusp-3Dsharing&amp;d=3DC=
wMFaQ&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3ybVka5=
SJOl76RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3DUTGtaoaFZgFyYb8eHhp2YE8F9ddj_g=
PZrcdAMKRGKK0&amp;s=3DRGFwTrPBgSIWNj0-NCA0r2nqe5zejReyedj-TrQY2gc&amp;e=3D=
" target=3D"_blank"> --><a href=3D"https://drive.google.com/file/" =
class=3D"">https://drive.google.com/file/</a><wbr =
class=3D"">d/0B412-7LDirTkTjBpZUV0ZVBGWnM<wbr =
class=3D"">/view?usp=3Dsharing<!-- </a> --></div>
<div class=3D""><br class=3D"">
</div>
<div class=3D"">Sincerely,</div>
<div style=3D"font-family: Tahoma; font-size: 13px;" class=3D"">
<div class=3D"">
<div class=3D"">
<div class=3D"">
<div class=3D"">
<div class=3D"">
<div style=3D"font-family:Tahoma;font-size:13px" class=3D""><font =
size=3D"3" face=3D"Verdana" class=3D""><br class=3D"">
</font></div>
<div class=3D""><font size=3D"2" face=3D"Arial" class=3D""><font =
class=3D"">Roland Zemla </font></font>
<div class=3D""><font size=3D"2" face=3D"Arial" class=3D"">MD/PhD =
Candidate</font></div>
<div class=3D""><font size=3D"2" face=3D"Arial" class=3D"">Basu =
Laboratory</font></div>
<div class=3D""><font size=3D"2" face=3D"Arial" class=3D"">NYU =
Neuroscience Institute</font></div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div><p class=3D""><br class=3D"">
------------------------------<wbr =
class=3D"">------------------------------<br class=3D"">
This email message, including any attachments, is for the sole use of =
the intended recipient(s) and may contain information that is =
proprietary, confidential, and exempt from disclosure under applicable =
law. Any unauthorized review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify =
the sender by return email and delete the original message. Please note, =
the recipient should check this email and any attachments for the =
presence of viruses. The organization accepts no
 liability for any damage caused by any virus transmitted by this =
email.<br class=3D"">
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr class=3D"">=3D=3D=3D</p>
</div>
</blockquote>
</div>
<br class=3D"">
</div>
</div>
</blockquote>
<blockquote type=3D"cite" class=3D"">
<div class=3D"">&lt;DemoMotionCorrection.py&gt;</div>
</blockquote>
<blockquote type=3D"cite" class=3D"">
<div class=3D"">&lt;demo_zemla.m&gt;</div>
</blockquote><p class=3D""><br class=3D"">
------------------------------<wbr =
class=3D"">------------------------------<br class=3D"">
This email message, including any attachments, is for the sole use of =
the intended recipient(s) and may contain information that is =
proprietary, confidential, and exempt from disclosure under applicable =
law. Any unauthorized review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify =
the sender by return email and delete the original message. Please note, =
the recipient should check this email and any attachments for the =
presence of viruses. The organization accepts no
 liability for any damage caused by any virus transmitted by this =
email.<br class=3D"">
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr class=3D"">=3D=3D=3D</p>
</div>
</blockquote>
</div>
</div>
</div>
</blockquote><p class=3D""><br class=3D"">
------------------------------<wbr =
class=3D"">------------------------------<br class=3D"">
This email message, including any attachments, is for the sole use of =
the intended recipient(s) and may contain information that is =
proprietary, confidential, and exempt from disclosure under applicable =
law. Any unauthorized review, use, disclosure, or distribution is =
prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the =
recipient should check this email and any attachments for the presence =
of viruses. The organization accepts no liability for any damage caused =
by any virus transmitted by this email.<br class=3D"">
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D<wbr class=3D"">=3D=3D=3D</p></div></div></div>

</blockquote></div><br class=3D""></div>
</div></blockquote></div><br class=3D""></body></html>=

--Apple-Mail=_EC756431-9D33-46BD-9B34-F6C9F7D8193D--

From dennisshasha@yahoo.com  Tue Sep 20 14:46:29 2016
Return-Path: <dennisshasha@yahoo.com>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u8KIkTDW013350
	for <shasha@mail.cims.nyu.edu>; Tue, 20 Sep 2016 14:46:29 -0400 (EDT)
Received: from nm37-vm8.bullet.mail.ne1.yahoo.com (nm37-vm8.bullet.mail.ne1.yahoo.com [98.138.229.136])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u8KIkLKo017301
	for <shasha@cims.nyu.edu>; Tue, 20 Sep 2016 14:46:21 -0400 (EDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s2048; t=1474397176; bh=1YcUvPlj6iwe4Yv+KbLCSUHfOJqXwF9bklq5eayQsYc=; h=Date:From:Reply-To:To:Cc:In-Reply-To:References:Subject:From:Subject; b=inVx7sxNYeaJKokJ9HIKtTZRREFzZF5YvQbvqZFXZgBtB/Ru/PH6zotBXqxURLPIuAb7LAnS5n7EM3dvwfXVSa25C8t2HtnzBA/9yVPU2f5Mo33ez8NG14nAkeTZEZvaO653DBUSnTJ3wo4GV0EAn2DD36jKNrvyvIFiZRHDnvI/UnXJaeDv1UoLdfhTBZ38YyPVxBZaDtFv1ht0JBydpMsK9Gd8Cu9+3itQFYeMv8fIB8NV6es/gbCho6BCXrZmuxskBwMkQDzyHpi/spwuHEkX6KSRbe2ZQgSe2/9bzTOjC8zCHPRv/7Wl6Y8AMQdL/geMgerxUsOQnUYupO1SnQ==
Received: from [127.0.0.1] by nm37.bullet.mail.ne1.yahoo.com with NNFMP; 20 Sep 2016 18:46:16 -0000
Received: from [98.138.226.176] by nm37.bullet.mail.ne1.yahoo.com with NNFMP; 20 Sep 2016 18:43:17 -0000
Received: from [98.138.89.244] by tm11.bullet.mail.ne1.yahoo.com with NNFMP; 20 Sep 2016 18:42:17 -0000
Received: from [127.0.0.1] by omp1058.mail.ne1.yahoo.com with NNFMP; 20 Sep 2016 18:42:17 -0000
X-Yahoo-Newman-Property: ymail-4
X-Yahoo-Newman-Id: 489544.97640.bm@omp1058.mail.ne1.yahoo.com
X-YMail-OSG: Au6JQzIVM1mQ_PayVYf7NevTjAsfs_IAlMeIs.7OX3hHVr5UZzYDSHxh39GfsXd
 .sOfjK1p40SQR02fwhFAb7HojXS9NxmCyiCY4TTyML51zfHKxl_w5FSIxYr7MWOmVoRWMMoadQ8z
 ySjRXtM.mVV7F3njvw61xcW0mUYIaWWZr4CSESJGp2GwQw20x3gsF_wQB5jNftyDC5zzNfoXDNoR
 SD39t3kHlB8mIrWJtR_dzJ4zH7Dkw5o8OL5Df43PImk43dIjslMJQ736lIaBn7XhCzFWpzXDX7Qn
 RJH62RRWbkwCKiLok3BzCwDmw1iSyOmqXeIGfw7M0zXopn6x166Q.JFabUYA9zPffQRejJsXHiQV
 aQqjjwTzdmE7JPNEiVMi5K1Gnm3.98ylhSCN6RHz0GpWrV4ULicoaPCLWtKpn2bZc_7FLCEpa8fv
 GsVTIZl1ztXodXSvAqyRG6O2H3ogoFLQ9JbewzuiUdDO8nH.F_AgV9nxGSX8uMESQst6b89zs7_g
 ZRXXzoySQ0hX6cXavqanrhE3Jk.i5ehiBsk_WSwqEcQ--
Received: from jws100187.mail.ne1.yahoo.com by sendmailws132.mail.ne1.yahoo.com; Tue, 20 Sep 2016 18:42:16 +0000; 1474396936.938
Date: Tue, 20 Sep 2016 18:42:16 +0000 (UTC)
From: dennis shasha <dennisshasha@yahoo.com>
Reply-To: dennis shasha <dennisshasha@yahoo.com>
To: "Basu, Jayeeta" <Jayeeta.Basu@nyumc.org>,
        Jonathan Kogan <jkogan18@cgps.org>
Cc: "Zemla, Roland" <Roland.Zemla@med.nyu.edu>,
        Dennis Shasha <shasha@cims.nyu.edu>
Message-ID: <1031679811.1851736.1474396936422@mail.yahoo.com>
In-Reply-To: <36b35b57-406e-46a4-b09a-4878c8e4b43c@MSGWBDCPVMHT62.nyumc.org>
References: <1703053827.155532.1471938041148.ref@mail.yahoo.com> <1703053827.155532.1471938041148@mail.yahoo.com> <131AC1E6-0759-498A-818A-95BA7C94DD8B@nyumc.org> <06daac37-8592-4b89-a67f-a0df0328611c@MSGWBDCPVMHT61.nyumc.org> <1347572069.656579.1471986427579@mail.yahoo.com> <ee95a64b-15b0-4d0d-81a8-f442c4f39ad0@MSGWBDCPVMHT64.nyumc.org> <CAG-=v_k2he08eeR37XLVLrRUcs7G024gWSkdX4hUnU5XcSChfg@mail.gmail.com> <CAG-=v_=-Ykm4i-uJLcEXOmiWDbSQmOLpFXH=3Uni7UEVkTw=Cg@mail.gmail.com> <36b35b57-406e-46a4-b09a-4878c8e4b43c@MSGWBDCPVMHT62.nyumc.org>
Subject: Re: current questions
MIME-Version: 1.0
Content-Type: multipart/alternative; 
	boundary="----=_Part_1851735_1713210451.1474396936382"
X-Greylist: Sender DNS name whitelisted, not delayed by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Tue, 20 Sep 2016 14:46:21 -0400 (EDT)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99
Status: R

------=_Part_1851735_1713210451.1474396936382
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Dear Colleagues,Well, it's been interesting. We have looked for a situation=
 in which sliding correlation should be stronger for=C2=A0close neurons tha=
n for far-away neurons. So far, distance doesn't seem to matter.Jonathan is=
 working on graphs in which neurons will be placed in the plane and highly =
correlated pairs of neuronswill be linked by edges (and highly negatively c=
orrelated pairs of neurons will be linked by edges of a different color).
We are taking small lags (up to 3 time points) into account.
Warmly,Dennis=20

    On Tuesday, September 20, 2016 12:40 PM, "Basu, Jayeeta" <Jayeeta.Basu@=
nyumc.org> wrote:
=20

  Dear Jonathan and DennisHow is the data analysis going?We are excited to =
see what the data shows.Best jayeeta

On Sep 7, 2016, at 11:18 AM, Jonathan Kogan <jkogan18@cgps.org> wrote:
Dear Dr. Basu,
Another question we have is is there a certain threshold below which we don=
't care about correlations and should skip over them? The reasoning behind =
this is we only want to find correlations when the neurons are stimulated. =
Could you please tell us what this threshold should be with the current dat=
a and with the absolute data when you send it. The threshold for the curren=
t data will be helpful because I'm writing the code for the analysis now in=
 preparation for the absolute data and need it for testing purposes.
Thanks,
Jonathan
On Fri, Sep 2, 2016 at 10:58 AM, Jonathan Kogan <jkogan18@cgps.org> wrote:

Dear Dr. Basu,
About the spatial information, would you be able to provide us with 3 dimen=
sional cartesian coordinates for each neuron to scale?
Best,
Jonathan

On Wed, Aug 24, 2016 at 10:10 AM, Basu, Jayeeta <Jayeeta.Basu@nyumc.org> wr=
ote:

Dear DennisRoland is preparing for his qualifying exam due at the end of th=
e month.=C2=A0I will discuss with him if he has the absolute fluorescence d=
ata at hand , otherwise we may need to rerun our analysis to extract this i=
nfo. This may take a little time.
I've responded to your other questions

Jayeeta=C2=A0
On Aug 23, 2016, at 5:07 PM, dennis shasha <dennisshasha@yahoo.com> wrote:


Dear Colleagues,Hmm. This dF/F issue brings up all kinds of new issues.If w=
e are going to look for correlations (please do answer those other question=
s in my email as well),I would really like to have absolute calcium levels,=
 because I think that's what we really want.But perhaps there is a calcium =
level below which we don't care what the value is. Is that the case?Thanks,=
Dennis

On Tuesday, August 23, 2016 2:19 PM, "Zemla, Roland" <Roland.Zemla@med.nyu.=
edu> wrote:


Dear Dennis and Jonathan,
a. Do you number frames starting at 0 or 1
The frame index begins at 1.
b. Some neurons have unchanging df/F (up to several decimal points). Is tha=
t a bug?
As far as I understand, the algorithm (http://www.cell.com/neuron/ab stract=
/S0896-6273%2815%2901084 -3)=C2=A0will smoothen and subtract the background=
 calcium trace and threshold the activity to create 'inferred' calcium even=
ts that the dF/F values reflect.=C2=A0 Events below threshold will have unc=
hanging dF/F.
A visual illustration of this transformation is shown in Figure 2B of that =
paper.
Best,
Roland ZemlaMD/PhD CandidateBasu LaboratoryNYU Neuroscience InstituteFrom: =
Basu, Jayeeta
Sent: Tuesday, August 23, 2016 7:12 AM
To: Zemla, Roland
Cc: dennisshasha@yahoo.com; jkogan18@cgps.org
Subject: Fwd: current questions

Dear Roland=C2=A0Could you please address the 2 questions Dennis has asked =
at the end of his enail?
Best=C2=A0
Jayeeta=C2=A0
Begin forwarded message:


From: dennis shasha <dennisshasha@yahoo.com>
Date: August 23, 2016 at 3:40:41 AM EDT
To: Jayeeta Basu <jayeeta.basu@nyumc.org>, Jonathan Kogan <jkogan18@cgps.or=
g>, Roland Zemla <roland.zemla@med.nyu.edu>, Dennis Shasha <dennisshasha@ya=
hoo.com>
Subject: current questions
Reply-To: dennis shasha <dennisshasha@yahoo.com>



Dear Colleagues,
Great. Thanks for your answers. Jonathan and I are going todo the following=
 analyses starting with absolute fluorescence (in the same format as before=
 please):
1. For control and the three stimuli, find the number of correlated ROIs ov=
er the first k% of frames following the stimulus onset, thenext k% etc. I'm=
 thinking that k% should be roughly 10%. Please tell me your views. We coul=
d also try various valuesand choose histograms.
Based on the kinetics of the calcium indicator the fastest event takes upto=
 75 ms to decay (3 frames). Multi spike events may last longer going upto 0=
.5 s. So that should be the time after stimulus for event detection=C2=A0

2. We will compare this number with the number of correlated ROIs for the c=
ontrol and use a p-value analysis to find significance.
Agreed
3. We will do the above for synchronous correlation, but we are thinking th=
at offsets (i.e. lagged correlations) might be interesting
too. If you agree, are there lags that you think would be particularly inte=
resting? We could try to discover good lags as well.
20 ms ; 70 ms; 125 ms are behaviorally relevant lags that could be interest=
ing. The onset of the calcium signal can be within the 5-10 ms range so we =
could detect the 20 ms lag possibly=C2=A0

4. The above analysis addresses the question "Does the stimulus cause unusu=
ally a lot of correlated reaction among a lot of ROIs and when following th=
e stimulus?" Equally interesting would be to identify which ROIs are correl=
ated and whether there isa spatial component. Do you have spatial informati=
on? If so, we could also look at the level of correlation as a function of =
distance.
We are interested in spatio-temporal patterns of activity . So yes a correl=
ation by distance or making a tonotopic map may be interesting. In terms of=
 behavior yes there may be certain neurons that respond to certain location=
s (of the animal) in space - mouse is running on a treadmill and we could p=
rovide the locomotive org signal to you that we have aquifer through a rota=
ry encoder on the wheel and RFID tag for track position.
5. Please let us know if you think we're missing something major.I think yo=
u've touched on most of the major analysis

Warm Regards,DennisP.S. Here are the additional questions that have come up=
 (mostly thanks to Jonathan):
a. Do you number frames starting at 0 or 1
b. Some neurons have unchanging df/F (up to several decimal points). Is tha=
t a bug?

------------------------------ ------------------------------
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D =3D=3D=3D



------------------------------ ------------------------------
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D =3D=3D=3D






------------------------------------------------------------
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D

  =20
------=_Part_1851735_1713210451.1474396936382
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<html><head></head><body><div style=3D"color:#000; background-color:#fff; f=
ont-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, =
sans-serif;font-size:16px"><div id=3D"yui_3_16_0_1_1474332034366_48055" dir=
=3D"ltr"><span id=3D"yui_3_16_0_1_1474332034366_48338">Dear Colleagues,</sp=
an></div><div id=3D"yui_3_16_0_1_1474332034366_48055" dir=3D"ltr"><span id=
=3D"yui_3_16_0_1_1474332034366_48341">Well, it's been interesting. We have =
looked for a situation in which sliding correlation should be stronger for&=
nbsp;</span></div><div id=3D"yui_3_16_0_1_1474332034366_48055" dir=3D"ltr">=
<span id=3D"yui_3_16_0_1_1474332034366_48238">close neurons than for far-aw=
ay neurons. So far, distance doesn't seem to matter.</span></div><div id=3D=
"yui_3_16_0_1_1474332034366_48055" dir=3D"ltr"><span id=3D"yui_3_16_0_1_147=
4332034366_48237">Jonathan is working on graphs in which neurons will be pl=
aced in the plane and highly correlated pairs of neurons</span></div><div i=
d=3D"yui_3_16_0_1_1474332034366_48055" dir=3D"ltr"><span id=3D"yui_3_16_0_1=
_1474332034366_48337">will be linked by edges (and highly negatively correl=
ated pairs of neurons will be linked by edges of a different color).</span>=
</div><div id=3D"yui_3_16_0_1_1474332034366_48055" dir=3D"ltr"><span><br></=
span></div><div id=3D"yui_3_16_0_1_1474332034366_48055" dir=3D"ltr"><span i=
d=3D"yui_3_16_0_1_1474332034366_48342">We are taking small lags (up to 3 ti=
me points) into account.</span></div><div id=3D"yui_3_16_0_1_1474332034366_=
48055" dir=3D"ltr"><span><br></span></div><div id=3D"yui_3_16_0_1_147433203=
4366_48055" dir=3D"ltr"><span>Warmly,</span></div><div id=3D"yui_3_16_0_1_1=
474332034366_48055" dir=3D"ltr"><span>Dennis</span></div> <div class=3D"qtd=
SeparateBR"><br><br></div><div class=3D"yahoo_quoted" style=3D"display: blo=
ck;"> <div style=3D"font-family: HelveticaNeue, Helvetica Neue, Helvetica, =
Arial, Lucida Grande, sans-serif; font-size: 16px;"> <div style=3D"font-fam=
ily: HelveticaNeue, Helvetica Neue, Helvetica, Arial, Lucida Grande, sans-s=
erif; font-size: 16px;"> <div dir=3D"ltr"><font size=3D"2" face=3D"Arial"> =
On Tuesday, September 20, 2016 12:40 PM, "Basu, Jayeeta" &lt;Jayeeta.Basu@n=
yumc.org&gt; wrote:<br></font></div>  <br><br> <div class=3D"y_msg_containe=
r"><div id=3D"yiv7145696084">

=20

<div>
Dear Jonathan and Dennis
<div class=3D"yiv7145696084">How is the data analysis going?</div>
<div class=3D"yiv7145696084">We are excited to see what the data shows.</di=
v>
<div class=3D"yiv7145696084">Best jayeeta</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
<div>
<blockquote type=3D"cite" class=3D"yiv7145696084">
<div class=3D"yiv7145696084">On Sep 7, 2016, at 11:18 AM, Jonathan Kogan &l=
t;<a rel=3D"nofollow" ymailto=3D"mailto:jkogan18@cgps.org" target=3D"_blank=
" href=3D"mailto:jkogan18@cgps.org" class=3D"yiv7145696084">jkogan18@cgps.o=
rg</a>&gt; wrote:</div>
<br class=3D"yiv7145696084Apple-interchange-newline">
<div class=3D"yiv7145696084">
<div dir=3D"ltr" class=3D"yiv7145696084">Dear Dr. Basu,
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084">Another question we have is is there a certain=
 threshold below which we don't care about correlations and should skip ove=
r them? The reasoning behind this is we only want to find correlations when=
 the neurons are stimulated. Could you please tell
 us what this threshold should be with the current data and with the absolu=
te data when you send it. The threshold for the current data will be helpfu=
l because I'm writing the code for the analysis now in preparation for the =
absolute data and need it for testing
 purposes.</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084">Thanks,</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084">Jonathan</div>
</div>
<div class=3D"yiv7145696084gmail_extra"><br class=3D"yiv7145696084">
<div class=3D"yiv7145696084gmail_quote">On Fri, Sep 2, 2016 at 10:58 AM, Jo=
nathan Kogan <span dir=3D"ltr" class=3D"yiv7145696084">
&lt;<a rel=3D"nofollow" ymailto=3D"mailto:jkogan18@cgps.org" target=3D"_bla=
nk" href=3D"mailto:jkogan18@cgps.org" class=3D"yiv7145696084">jkogan18@cgps=
.org</a>&gt;</span> wrote:<br class=3D"yiv7145696084">
<blockquote class=3D"yiv7145696084gmail_quote" style=3D"margin:0 0 0 .8ex;b=
order-left:1px #ccc solid;padding-left:1ex;">
<div dir=3D"ltr" class=3D"yiv7145696084">Dear Dr. Basu,
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084">About the spatial information, would you be ab=
le to provide us with 3 dimensional cartesian coordinates for each neuron t=
o scale?</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084">Best,</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084">Jonathan</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
</div>
<div class=3D"yiv7145696084HOEnZb">
<div class=3D"yiv7145696084h5">
<div class=3D"yiv7145696084gmail_extra"><br class=3D"yiv7145696084">
<div class=3D"yiv7145696084gmail_quote">On Wed, Aug 24, 2016 at 10:10 AM, B=
asu, Jayeeta <span dir=3D"ltr" class=3D"yiv7145696084">
&lt;<a rel=3D"nofollow" ymailto=3D"mailto:Jayeeta.Basu@nyumc.org" target=3D=
"_blank" href=3D"mailto:Jayeeta.Basu@nyumc.org" class=3D"yiv7145696084">Jay=
eeta.Basu@nyumc.org</a>&gt;</span> wrote:<br class=3D"yiv7145696084">
<blockquote class=3D"yiv7145696084gmail_quote" style=3D"margin:0 0 0 .8ex;b=
order-left:1px #ccc solid;padding-left:1ex;">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">Dear Dennis</div>
<div class=3D"yiv7145696084">Roland is preparing for his qualifying exam du=
e at the end of the month.&nbsp;</div>
<div class=3D"yiv7145696084">I will discuss with him if he has the absolute=
 fluorescence data at hand , otherwise we may need to rerun our analysis to=
 extract this info. This may take a little time.</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084">I've responded to your other questions<br clas=
s=3D"yiv7145696084">
<br class=3D"yiv7145696084">
Jayeeta&nbsp;</div>
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
On Aug 23, 2016, at 5:07 PM, dennis shasha &lt;<a rel=3D"nofollow" ymailto=
=3D"mailto:dennisshasha@yahoo.com" target=3D"_blank" href=3D"mailto:denniss=
hasha@yahoo.com" class=3D"yiv7145696084">dennisshasha@yahoo.com</a>&gt; wro=
te:<br class=3D"yiv7145696084">
<br class=3D"yiv7145696084">
</div>
</div>
</div>
<blockquote type=3D"cite" class=3D"yiv7145696084">
<div style=3D"background-color:rgb(255, 255, 255);font-family:HelveticaNeue=
, 'Helvetica Neue', Helvetica, Arial, 'Lucida Grande', sans-serif;font-size=
:16px;" class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084"><span class=3D"yiv7145696084">Dear Colleagues,=
</span></div>
<div class=3D"yiv7145696084"><span class=3D"yiv7145696084">Hmm. This dF/F i=
ssue brings up all kinds of new issues.</span></div>
<div class=3D"yiv7145696084"><span class=3D"yiv7145696084">If we are going =
to look for correlations (please do answer those other questions in my emai=
l as well),</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">I wo=
uld really like to have absolute calcium levels, because I think that's wha=
t we really want.</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">But =
perhaps there is a calcium level below which we don't care what the value i=
s. Is that the case?</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">Than=
ks,</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">Denn=
is</span></div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
<br class=3D"yiv7145696084">
</div>
</div>
</div>
<div style=3D"display:block;" class=3D"yiv7145696084">
<div style=3D"font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, =
Lucida Grande, sans-serif;font-size:16px;" class=3D"yiv7145696084">
<div style=3D"font-family:HelveticaNeue, Helvetica Neue, Helvetica, Arial, =
Lucida Grande, sans-serif;font-size:16px;" class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div dir=3D"ltr" class=3D"yiv7145696084"><font size=3D"2" face=3D"Arial" cl=
ass=3D"yiv7145696084">On Tuesday, August 23, 2016 2:19 PM, "Zemla, Roland" =
&lt;<a rel=3D"nofollow" ymailto=3D"mailto:Roland.Zemla@med.nyu.edu" target=
=3D"_blank" href=3D"mailto:Roland.Zemla@med.nyu.edu" class=3D"yiv7145696084=
">Roland.Zemla@med.nyu.edu</a>&gt; wrote:<br class=3D"yiv7145696084">
</font></div>
<br class=3D"yiv7145696084">
<br class=3D"yiv7145696084">
</div>
</div>
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div style=3D"direction:ltr;font-family:Tahoma;font-size:10pt;" class=3D"yi=
v7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv7145696084">Dear D=
ennis and Jonathan,</font>
<div class=3D"yiv7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv71=
45696084"><br class=3D"yiv7145696084">
</font></div>
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div dir=3D"ltr" style=3D"background-color:rgb(255,255,255);" class=3D"yiv7=
145696084"><span dir=3D"ltr" class=3D"yiv7145696084"><font face=3D"Arial" s=
ize=3D"2" class=3D"yiv7145696084">a. Do you number frames starting at 0 or =
1</font></span></div>
<div dir=3D"ltr" style=3D"background-color:rgb(255,255,255);" class=3D"yiv7=
145696084"><span dir=3D"ltr" class=3D"yiv7145696084"><font face=3D"Arial" s=
ize=3D"2" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</font></span></div>
<div dir=3D"ltr" style=3D"background-color:rgb(255,255,255);" class=3D"yiv7=
145696084"><span dir=3D"ltr" class=3D"yiv7145696084"><font face=3D"Arial" s=
ize=3D"2" class=3D"yiv7145696084">The frame index begins at 1.</font></span=
></div>
<div dir=3D"ltr" style=3D"background-color:rgb(255,255,255);" class=3D"yiv7=
145696084"><span dir=3D"ltr" class=3D"yiv7145696084"><font face=3D"Arial" s=
ize=3D"2" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</font></span></div>
<div dir=3D"ltr" style=3D"background-color:rgb(255,255,255);" class=3D"yiv7=
145696084"><span dir=3D"ltr" class=3D"yiv7145696084"><font face=3D"Arial" s=
ize=3D"2" class=3D"yiv7145696084">b. Some neurons have unchanging df/F (up =
to several decimal points). Is that a bug?</font></span></div>
<div class=3D"yiv7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv71=
45696084"><br class=3D"yiv7145696084">
</font></div>
<div class=3D"yiv7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv71=
45696084">As far as I understand, the algorithm (<a rel=3D"nofollow" target=
=3D"_blank" href=3D"https://urldefense.proofpoint.com/v2/url?u=3Dhttp-3A__w=
ww.cell.com_neuron_abstract_S0896-2D6273-252815-252901084-2D3&amp;d=3DDQMFa=
Q&amp;c=3Dj5oPpO0eBH1iio48DtsedbOBGmuw5jHLjgvtN2r4ehE&amp;r=3Dk3ybVka5SJOl7=
6RLO8VNy1GMcFoiF-463RN3SBJmj5o&amp;m=3D9zvUghBeVWJncChMX5vMiY1mrCXGe_b600co=
3XdkToQ&amp;s=3DPttkZvwUtkX1JTOmEbRFgRvo82DvJkB3PQIGjqC1rCc&amp;e=3D" class=
=3D"yiv7145696084">http://www.cell.com/neuron/ab stract/S0896-6273%2815%290=
1084 -3</a>)&nbsp;will
 smoothen and subtract the background calcium trace and threshold the activ=
ity to create 'inferred' calcium events that the dF/F values reflect.&nbsp;=
 Events below threshold will have unchanging dF/F.</font></div>
<div class=3D"yiv7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv71=
45696084"><br class=3D"yiv7145696084">
</font></div>
<div class=3D"yiv7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv71=
45696084">A visual illustration of this transformation is shown in Figure 2=
B of that paper.</font></div>
<div class=3D"yiv7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv71=
45696084"><br class=3D"yiv7145696084">
</font></div>
<div class=3D"yiv7145696084"><font face=3D"Arial" size=3D"2" class=3D"yiv71=
45696084">Best,</font></div>
<div class=3D"yiv7145696084">
<div style=3D"font-family:Tahoma;" class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div style=3D"font-family:Tahoma;font-size:13px;" class=3D"yiv7145696084"><=
font size=3D"3" face=3D"Verdana" class=3D"yiv7145696084"><br class=3D"yiv71=
45696084">
</font></div>
<div style=3D"font-size:13px;" class=3D"yiv7145696084"><font size=3D"2" fac=
e=3D"Arial" class=3D"yiv7145696084"><font class=3D"yiv7145696084">Roland Ze=
mla
</font></font>
<div class=3D"yiv7145696084"><font size=3D"2" face=3D"Arial" class=3D"yiv71=
45696084">MD/PhD Candidate</font></div>
<div class=3D"yiv7145696084"><font size=3D"2" face=3D"Arial" class=3D"yiv71=
45696084">Basu Laboratory</font></div>
<div class=3D"yiv7145696084"><font size=3D"2" face=3D"Arial" class=3D"yiv71=
45696084">NYU Neuroscience Institute</font></div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<div style=3D"font-family:'Times New Roman';font-size:16px;" class=3D"yiv71=
45696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<hr class=3D"yiv7145696084">
<div style=3D"direction:ltr;" class=3D"yiv7145696084"><font face=3D"Tahoma"=
 size=3D"2" class=3D"yiv7145696084"><b class=3D"yiv7145696084">From:</b> Ba=
su, Jayeeta<br class=3D"yiv7145696084">
<b class=3D"yiv7145696084">Sent:</b> Tuesday, August 23, 2016 7:12 AM<br cl=
ass=3D"yiv7145696084">
<b class=3D"yiv7145696084">To:</b> Zemla, Roland<br class=3D"yiv7145696084"=
>
<b class=3D"yiv7145696084">Cc:</b> <a rel=3D"nofollow" ymailto=3D"mailto:de=
nnisshasha@yahoo.com" target=3D"_blank" href=3D"mailto:dennisshasha@yahoo.c=
om" class=3D"yiv7145696084">
dennisshasha@yahoo.com</a>; <a rel=3D"nofollow" ymailto=3D"mailto:jkogan18@=
cgps.org" target=3D"_blank" href=3D"mailto:jkogan18@cgps.org" class=3D"yiv7=
145696084">
jkogan18@cgps.org</a><br class=3D"yiv7145696084">
<b class=3D"yiv7145696084">Subject:</b> Fwd: current questions<br class=3D"=
yiv7145696084">
</font><br class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084"></div>
</div>
</div>
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">Dear Roland&nbsp;</div>
<div class=3D"yiv7145696084">Could you please address the 2 questions Denni=
s has asked at the end of his enail?<br class=3D"yiv7145696084">
Best&nbsp;<br class=3D"yiv7145696084">
Jayeeta&nbsp;</div>
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
Begin forwarded message:<br class=3D"yiv7145696084">
<br class=3D"yiv7145696084">
</div>
<blockquote type=3D"cite" class=3D"yiv7145696084">
<div class=3D"yiv7145696084"><b class=3D"yiv7145696084">From:</b> dennis sh=
asha &lt;<a rel=3D"nofollow" ymailto=3D"mailto:dennisshasha@yahoo.com" targ=
et=3D"_blank" href=3D"mailto:dennisshasha@yahoo.com" class=3D"yiv7145696084=
">dennisshasha@yahoo.com</a>&gt;<br class=3D"yiv7145696084">
<b class=3D"yiv7145696084">Date:</b> August 23, 2016 at 3:40:41 AM EDT<br c=
lass=3D"yiv7145696084">
<b class=3D"yiv7145696084">To:</b> Jayeeta Basu &lt;<a rel=3D"nofollow" yma=
ilto=3D"mailto:jayeeta.basu@nyumc.org" target=3D"_blank" href=3D"mailto:jay=
eeta.basu@nyumc.org" class=3D"yiv7145696084">jayeeta.basu@nyumc.org</a>&gt;=
, Jonathan Kogan &lt;<a rel=3D"nofollow" ymailto=3D"mailto:jkogan18@cgps.or=
g" target=3D"_blank" href=3D"mailto:jkogan18@cgps.org" class=3D"yiv71456960=
84">jkogan18@cgps.org</a>&gt;,
 Roland Zemla &lt;<a rel=3D"nofollow" ymailto=3D"mailto:roland.zemla@med.ny=
u.edu" target=3D"_blank" href=3D"mailto:roland.zemla@med.nyu.edu" class=3D"=
yiv7145696084">roland.zemla@med.nyu.edu</a>&gt;, Dennis Shasha &lt;<a rel=
=3D"nofollow" ymailto=3D"mailto:dennisshasha@yahoo.com" target=3D"_blank" h=
ref=3D"mailto:dennisshasha@yahoo.com" class=3D"yiv7145696084">dennisshasha@=
yahoo.com</a>&gt;<br class=3D"yiv7145696084">
<b class=3D"yiv7145696084">Subject:</b> <b class=3D"yiv7145696084">current =
questions</b><br class=3D"yiv7145696084">
<b class=3D"yiv7145696084">Reply-To:</b> dennis shasha &lt;<a rel=3D"nofoll=
ow" ymailto=3D"mailto:dennisshasha@yahoo.com" target=3D"_blank" href=3D"mai=
lto:dennisshasha@yahoo.com" class=3D"yiv7145696084">dennisshasha@yahoo.com<=
/a>&gt;<br class=3D"yiv7145696084">
<br class=3D"yiv7145696084">
</div>
</blockquote>
</div>
</div>
<blockquote type=3D"cite" class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div style=3D"background-color:rgb(255, 255, 255);font-family:HelveticaNeue=
, 'Helvetica Neue', Helvetica, Arial, 'Lucida Grande', sans-serif;font-size=
:16px;" class=3D"yiv7145696084">
<div style=3D"display:table;width:955px;padding-top:12px;padding-left:0px;b=
ackground-color:inherit;" class=3D"yiv7145696084">
<div style=3D"display:table-cell;width:auto;word-wrap:break-word;" class=3D=
"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084">
<div class=3D"yiv7145696084"><span class=3D"yiv7145696084">Dear Colleagues,=
</span></div>
<div class=3D"yiv7145696084"><br clear=3D"none" class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084"><span class=3D"yiv7145696084">Great. Thanks fo=
r your answers. Jonathan and I are going to</span></div>
<div class=3D"yiv7145696084"><span class=3D"yiv7145696084">do the following=
 analyses starting with absolute fluorescence (in the same format as before=
 please):</span></div>
<div class=3D"yiv7145696084"><br clear=3D"none" class=3D"yiv7145696084">
</div>
<div class=3D"yiv7145696084"><span class=3D"yiv7145696084">1. For control a=
nd the three stimuli, find the number of correlated ROIs over the first k% =
of frames following the stimulus onset, the</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">next=
 k% etc. I'm thinking that k% should be roughly 10%. Please tell me your vi=
ews. We could also try various values</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">and =
choose histograms.</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084"><br =
class=3D"yiv7145696084">
</span></div>
</div>
</div>
<div dir=3D"ltr" class=3D"yiv7145696084">Based on the kinetics of the calci=
um indicator the fastest event takes upto 75 ms to decay (3 frames). Multi =
spike events may last longer going upto 0.5 s. So that should be the time a=
fter stimulus for event detection&nbsp;</div>
<span class=3D"yiv7145696084">
<div dir=3D"ltr" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">2. W=
e will compare this number with the number of correlated ROIs for the contr=
ol and use a p-value analysis to find significance.</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><br clear=3D"none" class=3D"yiv714=
5696084">
</div>
</span>
<div dir=3D"ltr" class=3D"yiv7145696084">Agreed</div>
<span class=3D"yiv7145696084">
<div dir=3D"ltr" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084">3. We will do the above for synchr=
onous correlation, but we are thinking that offsets (i.e. lagged correlatio=
ns) might be interesting<br clear=3D"none" class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">too.=
 If you agree, are there lags that you think would be particularly interest=
ing? We could try to discover good lags as well.</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><br clear=3D"none" class=3D"yiv714=
5696084">
</div>
</span>
<div dir=3D"ltr" class=3D"yiv7145696084">20 ms ; 70 ms; 125 ms are behavior=
ally relevant lags that could be interesting. The onset of the calcium sign=
al can be within the 5-10 ms range so we could detect the 20 ms lag possibl=
y&nbsp;</div>
<span class=3D"yiv7145696084">
<div dir=3D"ltr" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">4. T=
he above analysis addresses the question "Does the stimulus cause unusually=
 a lot of correlated reaction among a lot of ROIs and when following the st=
imulus?" Equally interesting would be to identify which ROIs are
 correlated and whether there is</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">a sp=
atial component. Do you have spatial information? If so, we could also look=
 at the level of correlation as a function of distance.</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084"><br =
class=3D"yiv7145696084">
</span></div>
</span>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">We a=
re interested in spatio-temporal patterns of activity . So yes a correlatio=
n by distance or making a tonotopic map may be interesting. In terms of beh=
avior yes there may be certain neurons that respond to certain locations
 (of the animal) in space - mouse is running on a treadmill and we could pr=
ovide the locomotive org signal to you that we have aquifer through a rotar=
y encoder on the wheel and RFID tag for track position.</span></div>
<span class=3D"yiv7145696084">
<div dir=3D"ltr" class=3D"yiv7145696084"><br clear=3D"none" class=3D"yiv714=
5696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">5. P=
lease let us know if you think we're missing something major.</span></div>
</span>
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084">I th=
ink you've touched on most of the major analysis</span></div>
<span class=3D"yiv7145696084">
<div dir=3D"ltr" class=3D"yiv7145696084"><span class=3D"yiv7145696084"><br =
class=3D"yiv7145696084">
</span></div>
<div dir=3D"ltr" class=3D"yiv7145696084"><br clear=3D"none" class=3D"yiv714=
5696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084">Warm Regards,</div>
<div dir=3D"ltr" class=3D"yiv7145696084">Dennis</div>
<div dir=3D"ltr" class=3D"yiv7145696084">P.S. Here are the additional quest=
ions that have come up (mostly thanks to Jonathan):</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084">a. Do you number frames starting a=
t 0 or 1</div>
<div dir=3D"ltr" class=3D"yiv7145696084"><br class=3D"yiv7145696084">
</div>
<div dir=3D"ltr" class=3D"yiv7145696084">b. Some neurons have unchanging df=
/F (up to several decimal points). Is that a bug?</div>
</span></div>
</div>
</div>
</div>
</div>
</div>
</div>
</blockquote>
</div>
</div>
</div>
</div>
<span class=3D"yiv7145696084">
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
------------------------------ ------------------------------<br class=3D"y=
iv7145696084">
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the re=
cipient should check this email and any attachments for the presence of vir=
uses. The organization accepts no
 liability for any damage caused by any virus transmitted by this email.<br=
 class=3D"yiv7145696084">
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D =3D=3D=3D</div>
</span></div>
</div>
<br class=3D"yiv7145696084">
<br class=3D"yiv7145696084">
</div>
</div>
</div>
</div>
</div>
</blockquote>
<span class=3D"yiv7145696084">
<div class=3D"yiv7145696084"><br class=3D"yiv7145696084">
------------------------------ ------------------------------<br class=3D"y=
iv7145696084">
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution
 is prohibited. If you have received this email in error please notify the =
sender by return email and delete the original message. Please note, the re=
cipient should check this email and any attachments for the presence of vir=
uses. The organization accepts no
 liability for any damage caused by any virus transmitted by this email.<br=
 class=3D"yiv7145696084">
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D =3D=3D=3D</div>
</span></div>
</blockquote>
</div>
<br class=3D"yiv7145696084">
</div>
</div>
</div>
</blockquote>
</div>
<br class=3D"yiv7145696084">
</div>
</div>
</blockquote>
</div>
<br class=3D"yiv7145696084">
</div>
<div><br>
------------------------------------------------------------<br>
This email message, including any attachments, is for the sole use of the i=
ntended recipient(s) and may contain information that is proprietary, confi=
dential, and exempt from disclosure under applicable law. Any unauthorized =
review, use, disclosure, or distribution is prohibited. If you have receive=
d this email in error please notify the sender by return email and delete t=
he original message. Please note, the recipient should check this email and=
 any attachments for the presence of viruses. The organization accepts no l=
iability for any damage caused by any virus transmitted by this email.<br>
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D</div></div>
</div><br><br></div>  </div> </div>  </div></div></body></html>
------=_Part_1851735_1713210451.1474396936382--


Dear Colleagues,
As I mentioned before, we are not finding much relationship between
distance and correlation though we are finding that many of the correlations
have to do with the center of the plane of neurons being tested.
So there is a spatial component.

The analyses here all have to do with a sliding window approach to correlation without
lag on raw fluorescence.
Each threshold in the name of the pdf refers to the correlation
threshold being tested. So .95 means a correlation threshold of 0.95 or greater
for each sliding window of 20 frames.
The positive and negative correlation values for
each combination are determined by the percent of correlations above the
positive threshold and the percent of correlations below the negative
threshold (e.g. >= 0.9 for positive and <= -0.95 for negative).
To show up on the graph, a pair needs to have a correlation above the threshold
(in the positive case) at for at least 35% of the sliding window frame pairs.
Similarly for the negative case at least 35% of the time below the negative correlation threshold.
In our analysis we detected only positively correlated pairs neurons. 

Please let us know if you have questions.

Warm Regards,

Jonathan and Dennis

P.S. Here is how to interpret filenames in case the above is not clear,
e.g. rawthresh75airpuff20Frame0LagConnections.pd
means positive correlations have to >= 0.7 for at least 35% of 20 long sliding windows
for the airpuff stimulus.


http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh70airpuff20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh70control20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh70Odor20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh70sound20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh75airpuff20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh75control20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh75Odor20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh75sound20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh80airpuff20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh80control20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh80Odor20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh80sound20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh90airpuff20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh90control20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh90Odor20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh90sound20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh95airpuff20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh95control20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh95Odor20Frame0LagConnections.pdf
http://cs.nyu.edu/cs/faculty/shasha/papers/rawthresh95sound20Frame0LagConnections.pdf
From shasha@cims.nyu.edu  Wed Sep 28 16:24:24 2016
Return-Path: <shasha@cims.nyu.edu>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u8SKOOwe009626
	for <shasha@mail.cims.nyu.edu>; Wed, 28 Sep 2016 16:24:24 -0400 (EDT)
Received: from linax2.cims.nyu.edu (linax2.cims.nyu.edu [128.122.49.117])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u8SKOJQm011955;
	Wed, 28 Sep 2016 16:24:19 -0400 (EDT)
Received: from linax2.cims.nyu.edu (localhost.localdomain [127.0.0.1])
	by linax2.cims.nyu.edu (8.14.4/8.14.4) with ESMTP id u8SKOIaA016959;
	Wed, 28 Sep 2016 16:24:18 -0400
Received: (from shasha@localhost)
	by linax2.cims.nyu.edu (8.14.4/8.14.4/Submit) id u8SKOIRs016958;
	Wed, 28 Sep 2016 16:24:18 -0400
From: Dennis Shasha <shasha@cims.nyu.edu>
Message-Id: <201609282024.u8SKOIRs016958@linax2.cims.nyu.edu>
Date: Wed, 28 Sep 2016 16:24:18 -0400
To: shasha@cs.nyu.edu, jgkogan@icloud.com, jkogan18@cgps.org
Subject: neural stuff
Cc: shasha@cs.nyu.edu
User-Agent: Heirloom mailx 12.4 7/29/08
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Wed, 28 Sep 2016 16:24:19 -0400 (EDT)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99



Dear Jonathan,
Somehow or another I mis-understood exactly what they want.
In fact what they want is somewhat simpler to compute, but has many
steps. It might be best for us to skype at some point to be
sure that this is clear to you, e.g. Saturday morning.
However, I will explain it as well as possible.
This will be both for raw fluorescence and for delta f/f_background.
Best,
Dennis


1. For each stimulus (air puff etc), take the 300 frames
before the stimulus and the 300 frames after.
Look at the average value before and after.
Find the p-value of the average_after - average_before using
a non-parametric shuffle test.
I will send you the python code for this.
We are interested in neurons that have a p-value <= 0.01 

2. Derivative question:
What percentage of cells respond to the odor vs. the air puff.

3. Derivative question:
For each neuron, which responds to which stimulus. We can use Sungear to
illustrate this. Contact Dennis McDaid: dm2936@nyu.edu

4. For each neuron that passes the p-value test, look for the first
peak and identify the time interval before the 
first peak when the neuron is between 5% and 95% of the peak.
Call that interval tau

5. For each stimulus what is the order of neuron firing.
Draw this as a directed graph where x --> y if  the end of the tau(x)
precedes the beginning of tau(y).

Warm Regards,
Dennis

From shasha@cims.nyu.edu  Wed Oct 12 17:39:19 2016
Return-Path: <shasha@cims.nyu.edu>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u9CLdJtg012402
	for <shasha@mail.cims.nyu.edu>; Wed, 12 Oct 2016 17:39:19 -0400 (EDT)
Received: from linax2.cims.nyu.edu (linax2.cims.nyu.edu [128.122.49.117])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id u9CLdHaK004662;
	Wed, 12 Oct 2016 17:39:17 -0400 (EDT)
Received: from linax2.cims.nyu.edu (localhost.localdomain [127.0.0.1])
	by linax2.cims.nyu.edu (8.14.4/8.14.4) with ESMTP id u9CLdH1m018839;
	Wed, 12 Oct 2016 17:39:17 -0400
Received: (from shasha@localhost)
	by linax2.cims.nyu.edu (8.14.4/8.14.4/Submit) id u9CLdH2q018838;
	Wed, 12 Oct 2016 17:39:17 -0400
From: Dennis Shasha <shasha@cims.nyu.edu>
Message-Id: <201610122139.u9CLdH2q018838@linax2.cims.nyu.edu>
Date: Wed, 12 Oct 2016 17:39:16 -0400
To: shasha@cims.nyu.edu, jkogan18@cgps.org
Subject: Re: Question About Tau
References: <CAG-=v_mEzfQ3voCu9UmB26O5HeSoOKPucg-A+Q-e0qvv0s_0BQ@mail.gmail.com>
In-Reply-To: <CAG-=v_mEzfQ3voCu9UmB26O5HeSoOKPucg-A+Q-e0qvv0s_0BQ@mail.gmail.com>
User-Agent: Heirloom mailx 12.4 7/29/08
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Wed, 12 Oct 2016 17:39:17 -0400 (EDT)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99
Status: O

... Dear Jonathan,

Jonathan Kogan <jkogan18@cgps.org> wrote:

> Hi Dennis,
>
> I'm writing the code to compute tau now and don't know what to do in
> certain scenarios. We are currently trying to find what neurons affect
> other neurons after the brain has been stimulated. So if one neuron spikes
> and then another one spikes entirely afterwards, it is affected by the
> other neuron. My question is what should I be using? The entire time series
> with before and after or just the after? Also, if I don't include the

... Just the after.

> before, what happens if the spike is entirely before the 300th frame? The
> stimulus would have had no affect so it wouldn't be important. It would
> just be coincidence that is happening right before the stimulus starts.

... I don't think it happens enough to worry about it.

>
> Also, if I do only analyze the after, what about the cases where it is
> spikes before and then is level in the after section. If we only analyze
> the after in those cases, there is no real spike or peak.

... Then it's not an interesting neuron:)

>
> Also, how do you suggest I actually find the peaks in general? There are
> the times when it goes up for 1-2 points and then back down and there are
> also ones where it goes up and down a lot and goes around the same height
> each time. Also, the scale for each neuron's spiking values is different.

... This is the question I asked and she sent me back a two sentence response.
I then asked for a more thorough response.

>
> What I think I'll do is define the peak as the highest point in the 300
> frames after. If the peak is at least 5 times higher than the minimum
> point, it will be considered an actual spike. I will then say the spike
> begins and ends when it gets below the mean value of the after frames on
> either side. Does that sound good to you as the preliminary tau analysis?

... Sure. Just to get the workflow working. Have you written to Dennis McDaid? Thanks, Dennis
>
> Thanks,
>
> Jonathan


From shasha@cims.nyu.edu  Wed Dec 21 12:04:38 2016
Return-Path: <shasha@cims.nyu.edu>
Received: from mx.cims.nyu.edu (mx.cims.nyu.edu [128.122.49.99])
	by mail.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id uBLH4cNm023485
	for <shasha@mail.cims.nyu.edu>; Wed, 21 Dec 2016 12:04:38 -0500 (EST)
Received: from CRUNCHY3.CIMS.NYU.EDU (crunchy3.cims.nyu.edu [172.22.49.57])
	by mx.cims.nyu.edu (8.15.1+Sun/8.15.1) with ESMTP id uBLH4Yi0016492;
	Wed, 21 Dec 2016 12:04:34 -0500 (EST)
Received: from CRUNCHY3.CIMS.NYU.EDU (localhost.localdomain [127.0.0.1])
	by CRUNCHY3.CIMS.NYU.EDU (8.14.4/8.14.4) with ESMTP id uBLH4YNC020445;
	Wed, 21 Dec 2016 12:04:34 -0500
Received: (from shasha@localhost)
	by CRUNCHY3.CIMS.NYU.EDU (8.14.4/8.14.4/Submit) id uBLH4YDb020441;
	Wed, 21 Dec 2016 12:04:34 -0500
From: Dennis Shasha <shasha@cims.nyu.edu>
Message-Id: <201612211704.uBLH4YDb020441@CRUNCHY3.CIMS.NYU.EDU>
Date: Wed, 21 Dec 2016 12:04:34 -0500
To: shasha@cims.nyu.edu, jkogan18@cgps.org
Subject: Re: Things to do
References: <2387e13f-caba-4627-9bda-c00e9a6eaf62@MSGWBDCPVMHT64.nyumc.org>
 <CAG-=v_=y1a_unAGY70xsvktzYQwn2LzmLnaApPt0gaeOCxuwyQ@mail.gmail.com>
 <201612181848.uBIImEEK029300@linax2.cims.nyu.edu>
 <CAG-=v_k8yYFJ0ZSvugSjUs7jZkYZRnz3r0tXbVv7n-yyFw_DMA@mail.gmail.com>
In-Reply-To: <CAG-=v_k8yYFJ0ZSvugSjUs7jZkYZRnz3r0tXbVv7n-yyFw_DMA@mail.gmail.com>
User-Agent: Heirloom mailx 12.4 7/29/08
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit
X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.4.1 (mx.cims.nyu.edu [128.122.49.99]); Wed, 21 Dec 2016 12:04:34 -0500 (EST)
X-Virus-Scanned: clamav-milter 0.97.4 at mx
X-Virus-Status: Clean
X-Scanned-By: MIMEDefang 2.73 on 128.122.49.99

... Dear Jonathan,

Jonathan Kogan <jkogan18@cgps.org> wrote:

> Hi Dennis,
>
> I just have a quick question about the pearson correlation analysis. First,
> the point of it is to find out what neurons are correlated under each
> stimulus. You said I should do a shuffle test to determine the significance
> of the correlation, but wouldn't it be more efficient if I just compared
> the correlation value to the entire distribution of correlation values? So
> I compare 1-2, 1-3, 1-4... 199-200 and then the ones with z scores above 2
> are considered correlated and are responding similarly. Also, when you
> mention delays, are those similar to what we had done in that original
> correlation analysis to determine if there was a delayed correlation? And
> if I'm going to compare the distribution, should I include the delayed ones
> in the distribution?

... For the delayed correlation. Take two time series T1, T2 corresponding to post-stimulus times.
Divide these into frames based on an input frame length (e.g. 8, 12, 20, whatever).
Each frame has a 1 or a 0 depending on whether that frame has an event or not.
For each delay (in units of frame length) such that the 1/0s in the
frames of T1 with that delay measure the Pearson correlation
with the 1/0s in the frame of T2 without a delay.
This will give a set of measured correlations corresponding to each delay (say a range of -3 to 
+ 3 in terms of frames).
Now, 10,000 times:
Shuffle the 1/0s in the frames of T2 and look at the correlation.
See where the various measured correlations fit in the sorted correlations.
If in the bottom 5 % (significantly negatively correlated) or top 5% (significantly positively correlated).

>
> Also, I should run the analysis on the data that I generated of groups of 8
> frames, right?
>
> Another thing I was thinking we could use correlation for is to determine
> similarities between effects of stimuli. So instead of doing it over many
> frames, we take the one that returned 1 value for each neuron. So if we
> found the stimulus affected neurons 1,4,7,13, and 18 in one stimulus and
> 2,4,7,16,18 in another, we could conclude that their stimuli affected them
> more similarly than another stimuli which stimulated neurons 8,9,10,11, and
> 12. Should I write code for this too?
>
> Best,
> Jonathan

... This is what Sungear will do. Best, Dennis
>
>
>
> On Sun, Dec 18, 2016 at 1:48 PM, Dennis Shasha <shasha@cims.nyu.edu> wrote:
>
> > ... Dear Jonathan,
> >
> > Jonathan Kogan <jkogan18@cgps.org> wrote:
> >
> > > Hi Dennis,
> > >
> > > I'm working on it and have a couple quick questions.
> > >
> > > 1. For the movies, do you want only the normal movies or also the movies
> > > that show unions and intersections over the different trials? Currently,
> > I
> > > only have included the normal movies.
> >
> > ... All of them.
> >
> > >
> > > 2. I'm a bit confused by this part of the description. What is the new
> > data
> > > without an external stimulus? Also, I understand the intersection part,
> > but
> > > what data am I doing it on? Do you want it me to do these analyses on the
> > > data where we don't break it up into groups where we just have the 1
> > before
> > > and after?
> >
> > ... I don't remember either what I meant here. In particular, I think I
> > meant
> > "with" even though I wrote "without". But it doesn't matter.
> > The point is, given a 1/0 matrix where each row represents a neuron,
> > compute a delayed correlation based on a specified delay, then do the
> > shuffle
> > test. To test it, do it on the data you already have with 1s and 0s.
> >
> > ... The second paragraph is different. For that, you take the analysis you
> > have now and decide using shuffle test which neurons respond for each
> > stimulus. Suppose there are n1 for stimulus 1, n2 for stimulus 2 etc.
> > Then you test the significance of the intersection as I specified.
> >
> > >
> > > "We will also look at new data without an external stimulus. He will give
> > > us a timestamp of each event 1/0. And then he’ll give us a delay and we
> > > will computer a delayed correlation between pairs of neurons to find
> > > putative causality.
> > > Then we will do a shuffle test on the 1s and 0s to see whether the
> > > correlation is significant. We will be using Pearson correlation. A
> > shuffle
> > > test just randomly shuffles the 0/1 vector and looks at the shuffled
> > > correlation to see whether the measured correlation is unlikely.
> > >
> > > We will also do a shuffle test to test the significance of the
> > > intersection. Here is how that goes: suppose there are stimuli A, B, C
> > and
> > > D having responses n1, n2, n3, n4 neurons respectively out of a total N.
> > > Suppose the total intersection is of size m. Is that intersection high or
> > > low?
> > > to these this, take an arbitrary n1 out N, n2 out of N, n3 out of N, n4
> > out
> > > of N and find the size of the intersection. Do this 10,000 times and see
> > > how often it is at least m."
> > >
> > > 3. About sungear. Their file lib/convertToJson.js has this at the
> > > beginning. I assume those are the files that are going to be created.
> > >
> > > var itemFile =
> > > '/Users/dennismcdaid/WebstormProjects/Sungear/data/items.json';
> > > var categoryFile =
> > > '/Users/dennismcdaid/WebstormProjects/Sungear/data/categories.json';
> > > var experimentFile =
> > > '/Users/dennismcdaid/WebstormProjects/Sungear/data/experiment.json';
> > >
> > > I don't know where they want me to specify the file that I want to input
> > > though. Also, from reading what they said, it sounds like they want me to
> > > just copy and paste the finished json information into that
> > routes/index.js
> > > file which currently has a small version of data in it. Is that actually
> > > how I need to do it?
> >
> > ... I don't know. Please ask Radhika and Dennis. It should use the format
> > I specified, but they should be helpful. Please copy me on the notes you
> > send them.
> >
> > >
> > > 4. I have attached what I have done so far in the zip file so you can see
> > > it. It does not include their sungear code because I didn't modify it at
> > > all. It does have a file called makeSungearData.py though that follows
> > your
> > > instructions on how to format the data to be ready to go into sungear.
> > The
> > > folder sungearData contains the formatted data. The readme is not well
> > > written yet and I will need to improve it later, but I think that it's
> > > enough for you to understand it so far.
> > >
> > >
> > > Best,
> > > Jonathan
> >
> > ... Ok, will look now. Warmly, Dennis
> > >
> >

