|
Post by hnorpois on Jun 23, 2014 15:02:28 GMT
Hello,
there are 2 ways to do a partioning of genomic variation by chromosome - separate and joint analysis. In Lee et al 2012 (Estimating the proportion of variation in susceptibility ...) it is mentioned : "... we performed two kinds of analyses: one in which the similarity matrix for each chromosome was fitted separately (22 analyses estimating one additive genetic variance component per analysis) and a second joint analysis in which the 22 similarity matrices were fitted simultaneously (estimating 22 additive genetic variance components in a single analysis)."
Young et al 2011 (Genome Partioning): "The joint analysis has the advantage of protecting against such inter-chromosomal correlations because the estimate of each is conditional on the other chromosomes in the model so that the estimates of variance explained by different chromosomes are independent of each other."
How is this done? Can anybody please comment on the differences between separated and joint analysis? And what is the corresponding gcta64 syntax?
Thanks Hermann Norpois
|
|
|
Post by Jian Yang on Jun 25, 2014 6:54:03 GMT
Please see Yang et al. 2011 Nat Genet. www.nature.com/ng/journal/v43/n6/abs/ng.823.htmlChromosomes are independent if there is no population structure but there will be a between-chromosome correlation at the presence of cryptic relatedness and/or population stratification. You can actually do a joint analysis of all 22 chromosomes using the --mgrm option in the REML analysis. # Multiple GRMs gcta64 --reml --mgrm multi_grm.txt --pheno test.phen --reml-no-lrt --out test_mgrm gcta64 --reml --mgrm multi_grm.txt --pheno test.phen --keep test.indi.list --reml-no-lrt --out test_mgrm
|
|
|
Post by hnorpois on Jun 25, 2014 13:03:11 GMT
Thanks for your reply. I apologise but ... I don't find the exact (mathematical) definition of joint analysis. The best I found was: "The joint analysis has the advantage of protecting against such inter-chromosomal correlations because the estimate of each is conditional on the other chromosomes in the model so that the estimates of variance explained by different chromosomes are independent of each other."
Why is the LRT skipped for multiple GRMs?
Thanks Hermann
|
|
|
Post by Jian Yang on Jun 26, 2014 11:52:07 GMT
See Yang et al. 2011 AJHG (page 2) and Yang et al. 2012 NG (Online Methods) for method details.
You can leave the --reml-lrt flag there but you need to specify which component you are going to test, e.g. --reml-lrt 2 means testing the second component (chr 2 in this case).
|
|