Gene mapping for MAGMA

151 views
Skip to first unread message

jelm...@gmail.com

unread,
Mar 3, 2020, 12:12:34 PM3/3/20
to FUMA GWAS users
Hi Kyoko,

This tool is great, thank you for all your work making this available!

Are there any plans in the future to use gene mapping such as eQTL or Chromatin Interactions to assign SNPs in MAGMA analyses? Alternatively, is there a way to output a file of SNPs and their annotated genes from all gene mapping methods that could be used as gene annotation file (*.genes.annot) in MAGMA run locally? 

Thanks,
Jeremy

Kyoko Watanabe

unread,
Mar 8, 2020, 5:53:47 PM3/8/20
to FUMA GWAS users
Hi Jeremy,

I'm sorry for the late response.

For now, there is no plan of using SNP2GENE gene mapping results for MAGMA gene analysis.
There are two main reasons.
First, MAGMA is designed to perform genome-wide analyses. If you use SNP2GENE gene mapping results, you are only using SNPs from GWAS loci which will limit the number of genes to analyze. If you use that results for gene-set or tissue specificity analyses, the results will be biased.
Second, SNP2GENE provides any overlapping annotations of SNPs in GWAS loci, however, as we all know, there are many false positives. For example, an overlap of eQTLs with GWAS loci could be due to LD. If you use all these SNPs to perform gene based test, the results will most likely be inflated.
For these reasons, I do not advice to directly use SNP2GENE mapped gene results for MAGMA analyses, you would need careful assessment.
Hope this helps.

Best,
Kyoko

jelm...@gmail.com

unread,
Mar 10, 2020, 6:18:23 PM3/10/20
to FUMA GWAS users
That makes sense, thank you for the explanation!
Reply all
Reply to author
Forward
0 new messages