colocalisation without software

[Daniel White]

Dear Marc,


No, no, and thrice No!

It is totally non scientific, completely subjective, and very often  
misleading
to just do 2 channel merges, of say green and red channel,
then just look for yellow.

Yes, you will see this non method all over cell biology journals,
but it flys in the face of all scientific logic.
It is very possible to get a merge image that appears to have yellow  
patches in it,
when in fact this is just random overlap and there is in fact no
statistically significant correlation between the objects in the 2  
channels.

If you want to even say anything, even qualitatively, about  
colocalisation,
then you must perform a channel correlation analysis, otherwise  
called colocalisation analysis.

There are very good imageJ plug ins to do this.

See
http://www.uhnres.utoronto.ca/facilities/wcif/imagej/colour_analysis.htm

You must also read the papers referenced in the plugins,
especially by Manders, and another by Costes.
The BioImageXD software project has implemented the same methods,
and is probably faster if you are doing many analyses.

You must understand the meaning of correlation coefficients,
Manders coefficients, how to interpret the channel 1 vs channel 2  
scatter plot
(where colocalised pixels will appear along a straight line),
how the auto threshold works,
and importantly, something which few people do, but everyone should
learn about and do:
  Fay, vanSeensel or best Costes method
statistical significance analysis of the colocalisation results
(these use randomisation of your image data and re analysis,
iterated may times, to see if the results from your data are any  
better than random or not.)

In order for these methods to work your image data must NOT be over  
exposed,
as if they are you have lost the most important part of the image  
data, and that will affect the analysis results in a bad way.

At the microscope use a rainbow or low high look up table to display  
the images you are taking, and keep the brightest pixels
well below the top of the range of the detector. You cant correct  
overexposed images later on with image processing.
If you clipped the data at acquisition, then its gone forever.

If I ever get a paper to review with a merge and no quantitative  
analysis, I will
send it straight back. It is quite simple to do, and adds much weight  
to the results.
There is very rich information in image data, and you should extract it!

cheers

Dan




Hi all,

It seems that in many papers from biologists or chemists, and i'm  
talking
high impact factors journals,  colocalisation of two elements is is  
often
assumed  by simple color superposition (ex: red and green fluoresce  
yellow
when colocalising), while microscopists (many physisists I suppose)  
seem to
need a more complex software-based confirmation.
Is it ok, when using high end equipment and corrected objectives  
(apochromat
with high NA for ex.), to assume colocalisation by color superposition,
especially when fluorophore are confined to small volume entities, like
lysosomes ?

Thanks

Marc

Dr. Daniel James White BSc. (Hons.) PhD
Bioimaging Coordinator
Nanoscience Centre and Department of Biological and Environmental  
Sciences
Division of Molecular Recognition
Ambiotica C242
PO Box 35
University of Jyväskylä
Jyväskylä
FIN 40014
Finland

[phone] (work)
[phone] (mobile)
http://www.bioimagexd.net
http://www.chalkie.org.uk
[hidden email]
[hidden email]