Different number of variants in summary statistics file and postgap results

Ensembl / postgap

Linking GWAS studies to genes through cis-regulatory datasets

Apache License 2.0

39 stars 21 forks source link

Different number of variants in summary statistics file and postgap results #174

Open 11Dmitriy11 opened 3 years ago

11Dmitriy11 commented 3 years ago

I found that postgap --summary_stats skips some variants from summary statistics file (variants that are parsed with --rsID and passes the threshold) and there are defferent number of variants in postgap result file and summary statiscits file. Is there a way to change the parameters so that result file contains all rsids (that can be successfully parsed) from summary statistics file.

dzerbino commented 3 years ago

Hello @11Dmitriy11 ,

There are indeed quite a few steps where SNPs are dropped, typically if they do not show up in the 1000 genomes genotypes. If you are willing to share with me the rsID of a SNP that was dropped, I can have a look at understanding why they dissappeared.

Regards,

Daniel

11Dmitriy11 commented 3 years ago

Hello @dzerbino, Here is list of rsids from summary file that a not present in gwas_snp and ld_snp_rsID columns, correspondingly. Also there are input summary file for postgap (UKBB phenotype 'never smoked'). P value threshold for postgap was 1e-4. 20116_0.gwas.imputed_v3.both_sexes_chr8_FILTERED.txt gwas_snp.txt ld_snp_rsID.txt I am interested in rs4246125 from intron region of PTK2 gene

11Dmitriy11 commented 3 years ago

Hello @dzerbino, Are there any updates?

dzerbino commented 3 years ago

Hello @11Dmitriy11 ,

No sorry, I've been snowed under the past few weeks, I'll try this weekend.

Regards,

Daniel