sethsh7/hla-prs-toolkit

Toolkit to aid generation of advanced Polygenic Risk Scores (PRS) including SNP-allele tagging and interaction models, especially in HLA. Developed and maintained by Seth Sharp ([email protected]).

hla-prs-toolkit's Issues

Typo in GRS67 mapping - 2 SNPs correspond to DQ92

In the file hla-prs-toolkit/Snplists/T1D_GRS67/mapping_TOPMED.txt, both rs28746898 and rs9405117 are marked as DQ92.

From the file T1D_GRS67_1000G_nopalin_pos_hg19.xlsx, it looks like rs9405117 should correspond to DQ93.

Missing strand for rs149663102

Add tiebreaker algorithm

Option for specifying ref alleles with third column

A future feature needs to be added that allows you to assign your score allele for DQ Haplotypes with a mapping, this is to avoid situations where one DQ type is >50% frequency and therefore will be wrongly called. This therefore only applies to cohorts heavily selected on certain DQ types.

Error in file validation code

rs2187668 position should be 6:32605884 not 6:4056277

In T1D_GRS10_1000G_nopalin_pos_hg19.xlsx
Might be the same typo in other data sheets, too.
Reference: https://www.ncbi.nlm.nih.gov/snp/rs2187668

Integration of arg parsing

Reported spreadsheet errors

(row 63) rs2111485 is on Chr2 not Chr1
(row 19) rs1281934 - major and minor may need swapping
(row 16) alleles not matching
(Row 6) two SNPs appear synonymous (possibly not an issue?)

matching on rsid

Hi I have imputed my data with Topmed imputation,
I wasnt provided rsid only position is this a problem?

Full GRS lists need MAF (requested)