ImageJ

Manders coefficient on timecourses

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

8 messages Options

Cath Heyward

Manders coefficient on timecourses

Hi all,

I'm looking for a way to measure Manders coefficients between two channels over time. So far I found a couple of plug-ins to measure Manders coefficients that work on stacks (e.g. JaCoP, Manders coefficients_class), but these seem to pool all the values into one instead of returning the individual values for each slice as I would need for a time course.

Is there a plug-in I haven't found that will do this for me?

I also tried writing a macro to use the plug-in on each slice individually (see below for code) but I get the same values for all the slices, so I assume this is still pooling together all the slices. I tried ticking the "Use current slice only" box but the macro recorder gives an error for the keyword "use" being used three times so I don't think the "use current slice only" command is being obeyed. Can anyone offer any suggestions for how to get round this?

Thanks in advance,

Cath

macro "Measure Manders Stack" {
frames=nSlices;
run("Clear Results");
for(i=0; i<frames; i++) {
currentslice=i+1;
setSlice(currentslice);
run("Manders Coefficients", "red=redtest.tif green=greentest.tif channel=[Red : Green] use=None use use exclude");
}
}

Jeremy Adler

Re: Manders coefficient on timecourses

A general request, when refering to Mander's coefficients it would be useful to make it clear which coefficients are being considered - an annoying feature of many papers.
The 1993 paper refers to M1 & M2, the Overlap coefficient and k1 & k2 - all could be called Mander's coefficients.
Referring to Mander's overlap coefficients is no better as this could cover M1 & M2 which cover the overlapping fraction each fluorophore and the Overlap coefficient.

M1 & M2 are clear and useful, the Overlap coefficient is far less useful than the Pearson correlation coeff (also features in the 1993 paper) while k1 & k2 derive from the Overlap coefficient and it is unclear, to me, what they really indicate.

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Cath Heyward [[hidden email]]
Sent: 20 March 2015 16:01
To: [hidden email]
Subject: Manders coefficient on timecourses

Hi all,

I'm looking for a way to measure Manders coefficients between two channels
over time. So far I found a couple of plug-ins to measure Manders
coefficients that work on stacks (e.g. JaCoP, Manders coefficients_class),
but these seem to pool all the values into one instead of returning the
individual values for each slice as I would need for a time course.

Is there a plug-in I haven't found that will do this for me?

I also tried writing a macro to use the plug-in on each slice individually
(see below for code) but I get the same values for all the slices, so I
assume this is still pooling together all the slices. I tried ticking the
"Use current slice only" box but the macro recorder gives an error for the
keyword "use" being used three times so I don't think the "use current slice
only" command is being obeyed. Can anyone offer any suggestions for how to
get round this?

Thanks in advance,

Cath

macro "Measure Manders Stack" {
frames=nSlices;
run("Clear Results");
for(i=0; i<frames; i++) {
currentslice=i+1;
setSlice(currentslice);
run("Manders Coefficients", "red=redtest.tif green=greentest.tif
channel=[Red : Green] use=None use use exclude");
}
}

--
View this message in context: http://imagej.1557.x6.nabble.com/Manders-coefficient-on-timecourses-tp5012083.html
Sent from the ImageJ mailing list archive at Nabble.com.

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

Cammer, Michael

Re: Manders coefficient on timecourses

Two weeks ago I did a comparison of JACoP with other Pearson's methods (my own, Volocity, and Imaris) and found that JACoP gave different answers for the Pearsons.

I find that people just toss images into these programs without any sort of characterization and often without understanding what colocalization means using different imaging techniques, magnifications, biological conditions, etc.

Are there any standard datasets out there for training people and testing software?

Thanks!

=========================================================================
Michael Cammer, Microscopy Core & Skirball Institute, NYU Langone Medical Center
Cell: 914-309-3270 Temporary location: SK2-7
http://ocs.med.nyu.edu/microscopy & http://microscopynotes.com/

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Jeremy Adler
Sent: Friday, March 20, 2015 12:24 PM
To: [hidden email]
Subject: Re: Manders coefficient on timecourses

A general request, when refering to Mander's coefficients it would be useful to make it clear which coefficients are being considered - an annoying feature of many papers.
The 1993 paper refers to M1 & M2, the Overlap coefficient and k1 & k2 - all could be called Mander's coefficients.
Referring to Mander's overlap coefficients is no better as this could cover M1 & M2 which cover the overlapping fraction each fluorophore and the Overlap coefficient.

M1 & M2 are clear and useful, the Overlap coefficient is far less useful than the Pearson correlation coeff (also features in the 1993 paper) while k1 & k2 derive from the Overlap coefficient and it is unclear, to me, what they really indicate.

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Cath Heyward [[hidden email]]
Sent: 20 March 2015 16:01
To: [hidden email]
Subject: Manders coefficient on timecourses

Hi all,

I'm looking for a way to measure Manders coefficients between two channels over time. So far I found a couple of plug-ins to measure Manders coefficients that work on stacks (e.g. JaCoP, Manders coefficients_class), but these seem to pool all the values into one instead of returning the individual values for each slice as I would need for a time course.

Is there a plug-in I haven't found that will do this for me?

I also tried writing a macro to use the plug-in on each slice individually (see below for code) but I get the same values for all the slices, so I assume this is still pooling together all the slices. I tried ticking the "Use current slice only" box but the macro recorder gives an error for the keyword "use" being used three times so I don't think the "use current slice only" command is being obeyed. Can anyone offer any suggestions for how to get round this?

Thanks in advance,

Cath

macro "Measure Manders Stack" {
frames=nSlices;
run("Clear Results");
for(i=0; i<frames; i++) {
currentslice=i+1;
setSlice(currentslice);
run("Manders Coefficients", "red=redtest.tif green=greentest.tif channel=[Red : Green] use=None use use exclude");
}
}

--
View this message in context: http://imagej.1557.x6.nabble.com/Manders-coefficient-on-timecourses-tp5012083.html
Sent from the ImageJ mailing list archive at Nabble.com.

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

jerie

Re: Manders coefficient on timecourses

On Mon, Mar 23, 2015 at 12:12 PM, Cammer, Michael <
[hidden email]> wrote:
---
Two weeks ago I did a comparison of JACoP with other Pearson's methods (my
own, Volocity, and Imaris) and found that JACoP gave different answers for
the Pearsons.

I find that people just toss images into these programs without any sort of
characterization and often without understanding what colocalization means
using different imaging techniques, magnifications, biological conditions,
etc.

Are there any standard datasets out there for training people and testing
software?
---

Michael,

very good topic, my impression is that careless use of coefficients has
made it very tedious to compare results.
For test sets, take a look at http://www.colocalization-benchmark.com/

Jens.

Dr. Jens Rietdorf

Visiting Scientist @ Center for Technological Development in Health (CDTS),
Oswaldo Cruz Foundation (Fiocruz), Ministry of Health, Rio de Janeiro,
Brazil.
http://br.linkedin.com/pub/jens-rietdorf/6/4a3/189/
Skype jens.rietdorf

On Mon, Mar 23, 2015 at 12:12 PM, Cammer, Michael <
[hidden email]> wrote:

> Two weeks ago I did a comparison of JACoP with other Pearson's methods (my
> own, Volocity, and Imaris) and found that JACoP gave different answers for
> the Pearsons.
>
> I find that people just toss images into these programs without any sort
> of characterization and often without understanding what colocalization
> means using different imaging techniques, magnifications, biological
> conditions, etc.
>
> Are there any standard datasets out there for training people and testing
> software?
>
> Thanks!
>
> =========================================================================
> Michael Cammer, Microscopy Core & Skirball Institute, NYU Langone Medical
> Center
> Cell: 914-309-3270 Temporary location:
> SK2-7
> http://ocs.med.nyu.edu/microscopy & http://microscopynotes.com/
>
> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
> Jeremy Adler
> Sent: Friday, March 20, 2015 12:24 PM
> To: [hidden email]
> Subject: Re: Manders coefficient on timecourses
>
> A general request, when refering to Mander's coefficients it would be
> useful to make it clear which coefficients are being considered - an
> annoying feature of many papers.
> The 1993 paper refers to M1 & M2, the Overlap coefficient and k1 & k2 -
> all could be called Mander's coefficients.
> Referring to Mander's overlap coefficients is no better as this could
> cover M1 & M2 which cover the overlapping fraction each fluorophore and the
> Overlap coefficient.
>
> M1 & M2 are clear and useful, the Overlap coefficient is far less useful
> than the Pearson correlation coeff (also features in the 1993 paper) while
> k1 & k2 derive from the Overlap coefficient and it is unclear, to me, what
> they really indicate.
>
>
> ________________________________________
> From: ImageJ Interest Group [[hidden email]] on behalf of Cath
> Heyward [[hidden email]]
> Sent: 20 March 2015 16:01
> To: [hidden email]
> Subject: Manders coefficient on timecourses
>
> Hi all,
>
> I'm looking for a way to measure Manders coefficients between two channels
> over time. So far I found a couple of plug-ins to measure Manders
> coefficients that work on stacks (e.g. JaCoP, Manders coefficients_class),
> but these seem to pool all the values into one instead of returning the
> individual values for each slice as I would need for a time course.
>
> Is there a plug-in I haven't found that will do this for me?
>
> I also tried writing a macro to use the plug-in on each slice individually
> (see below for code) but I get the same values for all the slices, so I
> assume this is still pooling together all the slices. I tried ticking the
> "Use current slice only" box but the macro recorder gives an error for the
> keyword "use" being used three times so I don't think the "use current
> slice only" command is being obeyed. Can anyone offer any suggestions for
> how to get round this?
>
> Thanks in advance,
>
> Cath
>
> macro "Measure Manders Stack" {
> frames=nSlices;
> run("Clear Results");
> for(i=0; i<frames; i++) {
> currentslice=i+1;
> setSlice(currentslice);
> run("Manders Coefficients", "red=redtest.tif green=greentest.tif
> channel=[Red : Green] use=None use use exclude");
> }
> }
>
>
>
>
> --
> View this message in context:
> http://imagej.1557.x6.nabble.com/Manders-coefficient-on-timecourses-tp5012083.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

Jens Rietdorf Visiting Scientist Fundação Oswaldo Cruz - Ministério da Saúde, Centro de Desenvolvimento Tecnológico em Saúde (CDTS), Rio de Janeiro, Brasil.

ctrueden

Re: Manders coefficient on timecourses

In reply to this post by Cammer, Michael

Hi Michael,

> Are there any standard datasets out there for training people and
> testing software?

The ImageJ wiki has a standard colocalization sample dataset linked from
the Colocalization Analysis page of the Cookbook documentation; see:

http://imagej.net/Colocalization#A_sample_dataset

Regards,
Curtis

On Mon, Mar 23, 2015 at 10:12 AM, Cammer, Michael <
[hidden email]> wrote:

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

Jeremy Adler

Re: Manders coefficient on timecourses

In reply to this post by Cammer, Michael

The lack of standard test datasets mostly arises because there are many coefficients in play and a test set of images that works for one coefficient may demonstrate the features of another poorly. For example an image intended to maximize the Manders M1 & M2 pair may have almost any correlation.

1) the simplest test is to take any image and use it for both channels. All correlation based methods should then produce a maximum value, usually 1.

2) the other end of the scale take two random images, these can easily be produced in imageJ - here the correlation should be zero, or close to zero. try File\New\Random

intermediate correlations can be generated by using method 1) and adding a random component to each using 2)

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Cammer, Michael [[hidden email]]
Sent: 23 March 2015 16:12
To: [hidden email]
Subject: Re: Manders coefficient on timecourses

Two weeks ago I did a comparison of JACoP with other Pearson's methods (my own, Volocity, and Imaris) and found that JACoP gave different answers for the Pearsons.

I find that people just toss images into these programs without any sort of characterization and often without understanding what colocalization means using different imaging techniques, magnifications, biological conditions, etc.

Are there any standard datasets out there for training people and testing software?

Thanks!

=========================================================================
Michael Cammer, Microscopy Core & Skirball Institute, NYU Langone Medical Center
Cell: 914-309-3270 Temporary location: SK2-7
http://ocs.med.nyu.edu/microscopy & http://microscopynotes.com/

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Jeremy Adler
Sent: Friday, March 20, 2015 12:24 PM
To: [hidden email]
Subject: Re: Manders coefficient on timecourses

A general request, when refering to Mander's coefficients it would be useful to make it clear which coefficients are being considered - an annoying feature of many papers.
The 1993 paper refers to M1 & M2, the Overlap coefficient and k1 & k2 - all could be called Mander's coefficients.
Referring to Mander's overlap coefficients is no better as this could cover M1 & M2 which cover the overlapping fraction each fluorophore and the Overlap coefficient.

M1 & M2 are clear and useful, the Overlap coefficient is far less useful than the Pearson correlation coeff (also features in the 1993 paper) while k1 & k2 derive from the Overlap coefficient and it is unclear, to me, what they really indicate.

________________________________________
From: ImageJ Interest Group [[hidden email]] on behalf of Cath Heyward [[hidden email]]
Sent: 20 March 2015 16:01
To: [hidden email]
Subject: Manders coefficient on timecourses

Hi all,

I'm looking for a way to measure Manders coefficients between two channels over time. So far I found a couple of plug-ins to measure Manders coefficients that work on stacks (e.g. JaCoP, Manders coefficients_class), but these seem to pool all the values into one instead of returning the individual values for each slice as I would need for a time course.

Is there a plug-in I haven't found that will do this for me?

I also tried writing a macro to use the plug-in on each slice individually (see below for code) but I get the same values for all the slices, so I assume this is still pooling together all the slices. I tried ticking the "Use current slice only" box but the macro recorder gives an error for the keyword "use" being used three times so I don't think the "use current slice only" command is being obeyed. Can anyone offer any suggestions for how to get round this?

Thanks in advance,

Cath

macro "Measure Manders Stack" {
frames=nSlices;
run("Clear Results");
for(i=0; i<frames; i++) {
currentslice=i+1;
setSlice(currentslice);
run("Manders Coefficients", "red=redtest.tif green=greentest.tif channel=[Red : Green] use=None use use exclude");
}
}

--
View this message in context: http://imagej.1557.x6.nabble.com/Manders-coefficient-on-timecourses-tp5012083.html
Sent from the ImageJ mailing list archive at Nabble.com.

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

Cath Heyward

Re: Manders coefficient on timecourses

In reply to this post by Cath Heyward

Thanks everyone for the input so far.
I was looking for a way to get the Mander's M1 and M2 coefficients out from the report from Coloc2 or JaCoP for a timecourse.
Is this not something many people use?

Thanks,

Cath

Olivier Burri

Re: Manders coefficient on timecourses

Hi Cath,

We had a similar need recently and made a small actionBar for that.

I have extracted the part that reads the Log Window and puts everything from JaCoP into a table and placed it in a public GIST

https://gist.github.com/lacan/c30610ef6aa016f8b078

Hope it is of use to you

Oli
BioImaging & Optics Platform

> -----Original Message-----
> From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of
> Cath Heyward
> Sent: Tuesday, March 24, 2015 9:03
> To: [hidden email]
> Subject: Re: Manders coefficient on timecourses
>
> Thanks everyone for the input so far.
> I was looking for a way to get the Mander's M1 and M2 coefficients out from
> the report from Coloc2 or JaCoP for a timecourse.
> Is this not something many people use?
>
> Thanks,
>
> Cath
>
>
>
> --
> View this message in context: http://imagej.1557.x6.nabble.com/Manders-
> coefficient-on-timecourses-tp5012083p5012125.html
> Sent from the ImageJ mailing list archive at Nabble.com.
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html