fitqtl effect estimates for X chromosome

23 views
Skip to first unread message

Thomas Blankers

unread,
Jan 25, 2023, 9:20:08 AM1/25/23
to R/qtl discussion
Dear community, Dear Karl,

it was not entirely clear to me whether  fitqtl estimates the effect of the X chromosome differently from the autosomes or whether the effect estimate for the X is corrected for the fact that one sex is hemizygous. So in our case we have only males phenotyped. The mating systen is XX X0, so males only have one copy of the X. The estimates for additive effects from fitqtl are roughly 1/2 the difference in phenotypic average between AA and BB genotypes, which corresponds to the effect of substituting one A allele for a B allele. For the X chromosome, the effect size should be not 1/2 but the whole difference between phenotypic average of A0 and B0. I could not quite figure out from the manual and the fitqtlengine code if this is indeed what happens or whether X chromosomal effects are biased downwards for the hemizygous sex in case of XX X0 mating systems. 

best

Thomas

Karl Broman

unread,
Jan 26, 2023, 9:45:55 AM1/26/23
to R/qtl discussion
Unfortunately, fitqtl is still not working, when it comes to estimated effects with a QTL on the X chromosome.
As it says in the help file:
The part to get estimated QTL effects is not complete for the case of the X chromosome and 4-way crosses. The values returned in these cases are based on a design matrix that is convenient for calculations but not easily interpreted.

The code to get estimated effects is in lines ~463-499 of fitqtl.R:

karl

Thomas Blankers

unread,
Feb 1, 2023, 7:56:13 AM2/1/23
to R/qtl discussion
Thanks! 

Op donderdag 26 januari 2023 om 15:45:55 UTC+1 schreef kbr...@gmail.com:
Reply all
Reply to author
Forward
0 new messages