snp2gene jobs questions

[tao pang]

Dear FUMA support team,

I am writing to report a potential issue with my SNP2GENE job (Job ID: 744845). After the analysis was completed, the "Genomic Loci" results only returned significant loci on two chromosomes, while a locus on chromosome 15 — which shows genome-wide significant associations in my input GWAS summary statistics — appears to be absent from the output.

I have verified that the chromosome 15 signal in my input data surpasses the genome-wide significance threshold (P < 5 × 10⁻⁸), and it is clearly visible in the Manhattan plot. However, it was not identified as a genomic risk locus in the SNP2GENE results table.

Could you please help me investigate why the chromosome 15 locus was not captured in the output? I would appreciate any guidance on whether this might be related to parameter settings (e.g., LD reference panel, maximum distance between lead SNPs, or SNP filtering criteria), or whether this could be a technical issue with the job run.

Thank you very much for your assistance.

Best regards,
Pang Tao

[Tanya Phung]

Hi Pang Tao, 

Since you clicked the option to keep the input files when you submitted your SNP2GENE job, I was able to check your input file on the FUMA server. 

Since you mention chromosome 15, I checked and there are indeed 12 variants where p values is less than 5e-8.

However, what is failing is the mismatching between the effect allele/non effect allele in your file versus the reference panel from the selected 1000G EUR population. 

For example, 

You have variant 15:22960447:C:A rs4457961

In FUMA database this variant is 15:22960447:G:T

Another example:
You have variant 15:23063749:G:A rs112228232

In FUMA database this variant is 15:23063749:C:T

You have 2 options: 

1. Make sure that the effect allele and non effect allele in your file is consistent with the reference panel. You can download the reference panel here: https://fuma.ctglab.nl/downloadPage. 
Scroll down to this section: 

2. You can remove the columns for effect allele and non effect allele before submitting your gwas sumstat to FUMA SNP2GENE. 

In this case, FUMA will look up the effect allele and non effect allele from the reference panel. 
For more information, you can check here: https://fuma-docs.readthedocs.io/en/latest/snp2gene/quick_start.html#examples-of-gwas-sumstat-for-snp2gene

Best,
Tanya

[tao pang]

Dear Tanya,

Thank you so much for your thorough check and clear explanation. I will follow your advice to align the alleles with the reference panel.

Much appreciated!

Best,
Pang Tao