Hello,
I have RNA-seq data that I would like to visualize via a GOterm heat map between time points. I have successfully been able to run DESeq2 on my transcriptome, but the output file with logpvalue/pvalue is my entire transcriptome, not compartmentalized by time point (0, 6, 12, 24, 48hr). I was able to make a heat map from this via time point but again, I have A LOT of genes so wanted to group by GOterm.
I have examined GO enrichment with programs like GOMWU, but again this examines the entire transcriptome, not each time point to compare between samples. Is there a way to use DESeq2 results and merge with GO terms to create a heat map?
I'm hoping my wording for this questions isn't too disorganized. Any suggestions on packages or programs that might help with my goal? I'm fairly new to all of this so apologies in advance.
There are choices to how you define the sample's performance in each GO term. I have some ideas for the way to do it. Try the package rgsepd which was designed for this purpose.
You might also try restating your goals. Saying you want to 'visualize via a GOterm heat map between time points' doesn't really explain exactly what you are after. One could assume that you want to find GO terms, for which the underlying set of genes appears to be perturbed in your experiment, and given those GO terms might want to make either individual heatmaps (by GO term), or a largish heatmap containing the genes in the top N GO terms (maybe separated into groups?).
But if you don't say exactly what you are after, you leave people to either guess at what you want or ignore your question entirely.
If I am right about what you want, things like goseq will get you halfway there (e.g., give you the GO terms that appear to be perturbed), and then you can make heatmaps using the genes in each term you care about.