search

CBIIT / LDlinkR

LDlinkR
56 stars 11 forks source link

LDlinkR: minor issue with LDexpress - column hg38 lists hg19 positions #18

Open timyers opened 2 years ago

timyers commented 2 years ago

Senf to LDlinkR maintainer email address from user ZEY, with minor edits:

I am contacting you in regards to your recently published R package LDlinkR. ...... However, I’d like to bring to your attention a minor issue that I have encountered with it that you might want to know about.

I was testing the LDexpress command as followed (requesting the SNPs to be annotated with hg38 positions in the output) LDexpress(snps = c("rs345", "rs456"), pop = c("CEU"), tissue = c("LIV", "TES"), r2d = "r2", r2d_threshold = "0.1", p_threshold = "0.1", win_size = "500000", genome_build = "grch38_high_coverage", token = "..." ) However, the output provided hg19 positions despite the column being named >Position_grch38_high_coverage<.

Query | RSID | Positiongrch38_high_coverage | R2 |   rs345 | rs756910 | chr13:32524656 | 0.473404255319149 |   rs345 | rs756911 | chr13:32525073 | 0.473404255319149 |   rs456 | rs17150305 | chr7:24891093 | 0.102998110552808 3 rows | 1-4 of 12 columns

The first SNV should be chr13:31950519 in hg38 build.

timyers commented 2 years ago

A minor typo was found in LDexpress() that resulted in the LDlink API always defaulting to genome build GRCh37, no matter which genome build was input by the user. The fix is now included in the development version on GitHub. It can be installed with:

remotes::install_github("CBIIT/LDlinkR")

This fix will be included with the next submission to CRAN, as soon as practical.