|
|
Post by elliew on Jul 13, 2016 9:38:51 GMT
Hi,
We're using GCTA cojo-slct to select signals from a meta-analysis, but were wondering whether GCTA will appropriately handle the X-chromosome? (using a population cohort of males & females for LD).
Many thanks,
Ellie
|
|
|
|
Post by nicolap on Jul 19, 2016 8:36:07 GMT
Hi,
I have tried to run the COJO analysis on chromosome X and it gives very weird results. I think one of the main reasons is that it seems that GCTA does not pick up the allelic frequencies from the reference set recognizing them as being 0. This makes p values become much smaller than the original ones.
Is there anything we should be doing to make it recognize the genotypes properly?
thanks a lot
Nicola
|
|
|
|
Post by Jian Yang on Aug 3, 2016 4:08:54 GMT
I haven't tested this but in principle it should work if you use females as the reference for LD estimation.
|
|
|
|
Post by nicolap on Aug 21, 2016 9:00:52 GMT
Hi,
I have tried this and it get very weird results, for example for one SNP p-value goes from 0.05 to 8.4x10-255 and there are different similar cases only on Chr X.
For the other chromosomes I do see some p-values improving but only sightly. These are drammatic differences.
Best
Nicola
|
|
