search

ht-diva / pqtl_pipeline_finemap

Fine mapping analysis within the pQTL pipeline project
MIT License
0 stars 0 forks source link

NLRP12 region #20

Closed dariushghasemi closed 2 weeks ago

dariushghasemi commented 3 weeks ago

Here we explain how we encountered the signals/loci falling in a specific region in CHR19 so-called NLRP12.

Adam said: "Looking at the results we found for Sun 2018 in INTERVAL, there were conditional lead SNPs that we attributed to the INTERVAL-specific NLRP12 locus ranging from 54,319,624 to 54,339,332 so quite a narrow (~20Kb) window."

He added: "I would suggest that we don’t remove all association results in a window, but go for your option of ‘Removing a locus only if the top snp is inside the problematic region’."

Therefore, we define the locus as chr19:54,300,000-54,360,000 (i.e. the previously seen NLRP12 region +/- about 20Kb).

To find the expanded coordinates in GRCh38, we used the original boundaries (chr19:54,319,624-54,339,332) and lifted the positions using this link, which led to this region in GRCh38: chr19:53,816,370-53,836,078.

After expanding the last region for +/- 20kb, we came up with chr19:53,796,000-53,856,000 and filter a locus only if the lead variant falling in this expanded window.

This filter was applied at two phases; one at the rule for collecting loci and the other at the COJO rule.

dariushghasemi commented 3 weeks ago

To filter the lead SNPs falling in NLRP12 region from the list of loci, we added a filter after merging loci and removing HLA region.

dariushghasemi commented 2 weeks ago

To filter NLRP12 region, user can only provide 1/0 standing for (Yes/No), to run or discard filter signals in this region and avoid running COJO.