Just wondering if someone can answer a question about the differences between the Venn Diagrams drawn in MapMan and Robin.

After doing the Robin analysis I get the 3 nice Venns describing the similarities between differences in my paired groups for "down", "up" and "total" -
Q1) what are the parameters for defining the "Significantly regulated genes" in the "total" diagram, I assume there must be some sort of filtering/threshold as it is evidently not the sum of the first two diagrams.

These Venns are nice however you do not get any information about gene annotation etc....

So onto MapMan to solve that, but;
Q2) Why when I use the outputs from Robin to draw Venns in MapMan they are totally different to the above, even after altering the threshold values.

Q3) Can I change the legend on the MapMan Venn diagrams

Could someone put me back on the right track, I've included the files if it helps

Martin.

Wheat affy arrays
MapMan 3.5.0

Hi Martin,

Unfortunately the Venn diagrams in Robin and MapMan filter the data based on
different characteristics which can be quite misleading. So to clarify that I'll answer
your questions in order:

Q1) Robin filters the genes based on the p-values assigned in the statistical analyses
of differential expression. If the threshold is not changed by the user this would be 0.05
or lower. MapMan on the other side filters based on log2 fold change and ignores the
p-values (or 1/0 flags) for now. Concerning the annotation: The Venn diagrams that
Robin generates are just static overview images and hence not interactively providing
annotation information. The data that is used to generate them, however, is supplies in
the result files so you can have a look into these to get to the annotation (provided you
chose to annotate the results at the end of the Robin workflow)

Q2) see above - MapMan and Robin filter the data differently

Q3) I fear you can't

I hope this answers your questions - don't hesitate to aks again if you have more
questions.

cheers,
Marc

Hi Marc,

Thanks for the info it's becoming clearer.

However just one last question, and sorry if it's a little basic but hey we're all allowed one stupid question a day

So if I want to draw a Venn in MapMan from the "full_table" or "top100table" should choose the logFC data? (seems to give sensible results both on numbers and regulated gene families, but thought I had better check!)

Much appreciated and many thanks for building Robin it's really helping our group to avoid the leap into R,

Martin.

Hi Martin,

and don't worry; There's no stupid questions only stupid answers. To answer your question:
Ideally, you would simply use the main results file that Robin generates. It contains a logFC and
a significance flag column for each of the contrasts you defined - so yes, you would use the
logFC data.

cheers,
Marc