GigaMUGA consesnus SNPs

Richard Radcliffe

unread,

Mar 10, 2021, 11:37:50 AM3/10/21

to R/qtl2 discussion

Karl-

The consensus allele code file that you present for the GigaMUGA array (GM_allelecodes.csv) contains 112,913 markers, ~30k markers fewer than what's on the array. I can't seem to find anything that explains why these markers were removed. Any insight? Thanks.

Karl Broman

unread,

Mar 10, 2021, 11:45:24 AM3/10/21

to R/qtl2 discussion

Construction of my annotation files explained at https://kbroman.org/MUGAarrays/new_annotations.html

Construction of the founder genotype files are in this script: https://figshare.com/articles/dataset/parse_muga_R/5405260

I think the main reduction in markers was because they were not polymorphic in the DO founders.

karl

Richard Radcliffe

unread,

Mar 10, 2021, 3:59:47 PM3/10/21

to R/qtl2 discussion

Thanks for the info. I'm working with the HS/Npt mice which have a different but overlapping set of founders from the DO (HS/Npt: AKR/J, BALB/cJ, C3H/HeJ, DBA/2J, LP/J, C57BL/6J, A/J, CBA). For what it's worth, I just checked those 30k excluded markers and 3,175 are informative in the HS/Npt. Among the set of 112,913 markers, 51,658 are uninformative in the HS/Npt.