Visualization of fold changes of candidate genes as heatmap across multiple independent experiments

Question

[VBer]

Hello.

I have a candidate set of genes whose expression I would like to visualize in heatmap format.

As I have several transcriptome/proteome studies, I cannot represent them all as a single heatmap, as log2FC is relative.

Is there any global variable I can calculate to compare/visualize the fold changes of these genes? My goal is to look at a particular gene across several studies and see if its expression is consistent.

All the studies, although from different independent experiments are from the same tissue samples.

Thanks!

Answer 1

[jared.andrews07]

The most simple solution may be to generate each of the individual heatmaps and just edit them together in an aesthetically pleasing way via Illustrator or such. This has the benefit of being explicit about the source of each dataset, though it may be rather tedious to create. And if you have a large number of datasets, may be impractical. Showing it for a handful of them and sticking the rest in a supplemental figure may be a feasible approach.

Another option would be to do something like pre-ranked GSEA for your genes of interest using the results of each separate DGE analysis to show that they are enriched in the up/downregulated genes for each condition in each dataset as you'd expect. This would allow generation of a straightforward summary figure, but unfortunately obfuscates potential outliers/genes that don't agree in all datasets.

[VBer]

Thanks for your answer! I think I'll go with the GSEA option! I'll update later on how that worked out for me.