I am working on a bird species with more than 30 chromosomes.
I can happily input .tped and .tped files from Plink and run an analysis in GenABEL.
However, in Plink, my sex chromosome is coded 35, and I want GenABEL to know it is not autosomal.
I can recode this as X using the recodeChromosome function.
Recoded chromosome for 28650 SNPs ( 35 -> X )
[Note that birds are actually ZZ male / ZW female, so the 'X' here corresponds to the 'Z' and I have swapped the male and females around in the sex column, such that males are 0 and females are 1, to account for females being heterogametic]
However, after doing this, I had a problem with check.marker, such that it now hangs at the 'Passed: X markers and Y people' stage, before doing the quality control.
This doesn't happen when I run the same file but with the sex chromosome still coded as an integer.
I wonder whether the problem is because I still have heterozygous male (really female) genotypes.
However, it seems that Xfix() should only be run to correct these AFTER running check.marker.
Anyone else encoutnered this problem/got a solution?
Relevant commands and output below.
Code: Select all
Excluding people/markers with extremely low call rate...
494705 markers and 491 people in total
0 people excluded because of call rate < 0.1
0 markers excluded because of call rate < 0.1
Passed: 494705 markers and 491 people
Running sex chromosome checks...
53272 heterozygous X-linked male genotypes found
402 X-linked markers are likely to be autosomal (odds > 1000 )
0 male are likely to be female (odds > 1000 )
0 female are likely to be male (odds > 1000 )
0 people have intermediate X-chromosome inbreeding (0.5 > F > 0.5)
If these people/markers are removed, 32844 heterozygous male genotypes are left
... these will be considered missing in analysis.
... Use Xfix() to fix these problems.
Passed: 494303 markers and 491 people