peak2gene using geneScore

[Brawni]

Hello,

Im wondering whether running the peak2gene analysis correlating distal peaks to genescore instead of gene expression would be still interesting once the genescore has been recomputed to sum up only fragments in promoter and or overlapping gene body? Is there an easy way to do this?

Thank you!

[rcorces]

I've never done this and there are obvious reasons why this could be problematic but I think you should be able to do what you want just by setting useMatrix to GeneScoreMatrix.
Related, the parameter documention for addPeak2GeneLinks() was incomplete and I updated it via 2aabfc3

[Brawni]

I see thanks. But if you recompute the genescore to consider only fragments in promoter / gene body and then run the analysis would that still be problematic? and if yes why?

[rcorces]

I havent thought this through completely but I think there are certain precautions you would need to take. For example, it could be problematic if your regulatory element is an intronic enhancer, this will contribute to the gene score.

[Brawni]

So if I only wanted to use fragments mapping to the promoter region of the gene as a proxy for gene expression, what should it be the correct settings in the addGeneScoreMatrix ?

[rcorces]

useTSS = TRUE will disable gene body extension.
I think there is actually a bug in the code (or documentation) for this use case though.
You'll have to also set extendTSS=TRUE and this will extend the TSS region on either side based on geneUpstream and geneDownstream as shown here:

ArchR/R/MatrixGeneScores.R

Line 224 in 968e442

if(extendTSS){

However, the documentation would make you believe that extendUpstream and extendDownstream would be the proper params for this.

[Brawni]

ok thanks a lot! I will give it a go!