Posted by
Jeremy Adler-4 on
URL: http://confocal-microscopy-list.275.s1.nabble.com/Colocalization-analysis-in-Fiji-ImageJ-Coloc-2-plugin-updated-with-several-fixes-tp6955853p6962540.html
*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy*****
It is hard enough to get scientists to make and journals to require
measurements of colocalization without (George) teasing everyone by
announcing colocalization is dead.
Hi resolution imaging has not reached molecular resolution, but
increasing resolution creates some intriguing problems. a) most
molecules are not inherently fluorescent so we image and localise a
fluorescent tag that is offset from the target b) stochastic methods
have their own noise - the probability of a molecule being imaged or
even reimaged. Molecular Poisson noise will impair nearest neighbour
analyses as will the presence of untagged molecules.
But back in the imaging world most of us inhabit we should appreciate
that colocalization measurements divide into two categories, those
that characterise cooccurence (M1 and M2) and those that characterize
correlation (Pearson & Spearman) among the fluorescence that show
cooccurence. There is third category of coefficients that are
inherently confusing and perhaps a fourth for coefficients those are
incorrectly applied.
Adler J., Parmryd I. Cytometry A. 2010, 77(8), 733-42.
In answer to Dan's question including the Spearman correlation
coefficient is worthwhile, it is unaffected by non linearities and is
less senstive to outliers than the Pearson correlation coefficient.
Quoting George McNamara <
[hidden email]>:
> *****
> To join, leave or search the confocal microscopy listserv, go to:
>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy> *****
>
> Hi Dan,
>
> Too bad red + green = yellow pixel colocalization and its many
> coefficients is dead. See PubMed 21809413 - do you have nearest
> neighbor analysis in yet/
>
> George
>
> On 11/2/2011 10:54 AM, Daniel James White wrote:
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>>
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>> *****
>>
>> Dear colocalization analysis fans.
>>
>> Thanks the keen eyes and astute minds of users
>> (Thanks to all of you!!!)
>> of the new working prototype
>> Coloc_2 plugin for Fiji (is just imageJ - batteries included)
>> we have identified and fixed an number of bugs and design problems
>> in the code:
>>
>> 1) A problem in the design of how regions of interest/mask were implemented
>> caused a bug in the thresholded Manders coefficient calculation
>> that gave values greater than 1.
>> This should now be fixed. Please test it as tell us if you feel it
>> does the right thing.
>> (JACoP might give different numbers due to a different
>> interpretation of the maths in the original Costes paper.
>> After asking Dr. Costes, we believe Coloc_2 is correct)
>>
>> 2) Some further fixes were made in how masks/ROIs are handled for the
>> Costes statistical significance randomization test.
>>
>> 3) We added test code that uses Gaussian shaped spot images of
>> various patterns
>> as per Manders' original paper, in order to recapitulate his tests
>> and to identify
>> problems in our implementations of algorithms which calculate
>> Pearsons and Manders coefficients etc.
>>
>> There are also various little tweaks and redesigns,
>> that should make it easier for any programmer to read, follow and understand
>> and then extend the code with new functionality
>> (Spearman correlation anyone? Objects based analysis like in JACoP
>> - anyone?)
>>
>> Look in the GIT log to see all the code revision commit messages
>> from Tom and Johannes.
>> The ideas/design document is still viewable at
>>
https://docs.google.com/document/d/1bEJyXdGyKx4J_G6WlsABpcl3cn-kI5CQs57qU-3Gx7Y/edit>>
>>
>> Thanks to all involved,
>> and please let us know how you would like the user interface and
>> the standardized output PDF to look.
>>
>> happy colocalizing... update Fiji and you will get the new version
>> of Coloc_2.
>>
>> cheers
>>
>> Dan
>>
>>
>>
>> Dr. Daniel James White BSc. (Hons.) PhD
>>
>> Leader - Image Processing Facility,
>> Senior Microscopist,
>> Light Microscopy Facility.
>>
>> Max Planck Institute of Molecular Cell Biology and Genetics
>> Pfotenhauerstrasse 108
>> 01307 DRESDEN
>> Germany
>>
>> +49 (0)15114966933 (German Mobile)
>> +49 (0)351 210 2627 (Work phone at MPI-CBG)
>> +49 (0)351 210 1078 (Fax MPI-CBG LMF)
>> chalkie666 Skype
>>
http://www.bioimagexd.net BioImageXD
>>
http://fiji.sc Fiji - is just ImageJ (Batteries Included)
>>
http://www.chalkie.org.uk Dan's Homepages
>>
https://ifn.mpi-cbg.de Biopolis Dresden Imaging Platform (BioDIP)
>> dan (at) chalkie.org.uk
>> ( white (at) mpi-cbg.de )
>>
>>
>
>
> --
>
>
> George McNamara, PhD
> Analytical Imaging Core Facility
> University of Miami
>
Jeremy Adler
IGP
Rudbeckslaboratoriet
Daghammersköljdsväg 20
751 85 Uppsala
Sweden
0046 (0)18 471 4607