Error on P-value threshold when using SNP2GENE (maximum lead SNP P-value is 1e-5)

[Roy Costilla]

Dear Kyoko,

I'm trying to map a specific region using SNP2GENE in FUMA and for my GWAS this region didn't reach very low p-values. As a consequence, I'm getting the error "maximum lead SNP P-value is 1e-5" (see below). I understand that standard genome-wide significance thresholds are important but I think less stringent thresholds for specific regions should also be allowed in FUMA.  Am I missing something important or this just a bug that can be corrected? 

Many thanks,

Roy

2. Parameters for lead SNPs and candidate SNPs identification

Minimum P-value of lead SNPs (<)

The maximum lead SNP P-value is 1e-5.

[Kyoko Watanabe]

Hi Roy,

I completely understand, when you don't have genome-wide significant hits, you might want to annotate top hits regardless of P-value.

The reason why I'm setting 1e-5 as the maximum P-value is that, if I don't limit anything, people start using P=1 meaning try to annotate every SNP in the input file.

But FUMA is primary aimed to annotated GWAS risk loci, but not to provide genome-wide annotation.

And if every one start annotating all SNPs, server will crush.

I am sorry but I cannot change this setting just for you.

But here is possible work around.

You can provide pre-defined lead SNPs. Those SNPs will be considered lead SNPs regardless of their P-value.

So if you can clump by yourself and submit list of lead SNPs together with your summary statistics, you can annotated those loci.

Please let me now if you need any help for that.

Best,

Kyoko