Clarification on “protein-coding genes only” option

[jiwon park]

Dear FUMA team,

I have a question regarding the “Gene types” setting (where “Protein coding” is selected) in FUMA.

In my analysis, I selected “Protein coding” in the Gene types section, expecting that the mapped gene annotation results would be restricted to protein-coding genes. However, I still observe that non-coding genes are included in the annotation outputs.

This led to some confusion about how this option is applied.

Specifically, I would like to clarify:

• Is the Gene types selection (e.g., “Protein coding”) intended to filter the SNP-to-gene annotation results themselves?
• Or is it only applied to downstream analyses (e.g., gene-set or MAGMA analyses), while the annotation step still includes all gene types?

If the annotation results are not filtered by this option, would it be appropriate for users to manually restrict to protein-coding genes when reporting results?

Thank you very much for your clarification.

Best regards,
Jiwon Park

[Tanya Phung]

Hi Jiwon, 

Could you provide more information on: 

• When you said: "I still observe that non-coding genes are included in the annotation outputs.", what is the jobID? Which non-coding genes did you observe and from which output? 

To answer your question, as far as I am aware, the selection of gene type applies to the gene mapping step. In this section of the code https://github.com/vufuma/FUMA-webapp/blob/master/scripts/genemap/geneMap.R#L87-L90

if(genetype!="all"){
genetype <- unique(unlist(strsplit(genetype, ":")))
ENSG <- ENSG[ENSG$gene_biotype %in% genetype,]
}

ENSG is filtered such that it contains the selected gene type before it does additional downstream analyses. 

However, bugs are common so there may be a bug somewhere. To help with understanding what is going on, I can take a closer look once you send me the requested information above. 

Best,
Tanya