I used the GATK pipeline to call SNPs on males and females using RNA-Seq data. But the males have a higher read count (~43-46M reads) than the females (~40-42M reads). This causes SNP counts to be higher for males and lower for females. This is strange because I am interested in X-chromosome and hence the question "How to control for the number of reads when calling SNPs using RNA-Seq?"
4
-
$\begingroup$ I think there is no way to do this, the ultimate solution is to get more data. If you have more reads, you naturally have more SNPs called. You can find the intersection of males and females if you want to analyze the two together. $\endgroup$– Phoenix MuCommented Dec 18, 2020 at 23:27
-
$\begingroup$ would it be acceptable to subsample all the reads at 40M reads with the same seed and then do the GATK pipeline? I am thinking of seqtk to subsample. $\endgroup$– BalanCommented Dec 19, 2020 at 0:56
-
$\begingroup$ That is not even 10% difference, plus you would need to check whether the one group has the same number of reads aligning to the exome. It could be that one group has more gDNA contamination, or rRNA, or whatever, and naive per-million scaling would fail. I am not sure though that all is necessary here, just run the pipeline RNA-seq is anyway suboptimal for variant calling and you will need to confirm interesting findings with an indepedent approach, so don't overthink it at this stage, run the pipeline and see whether you even find something interesting at all. $\endgroup$– user3051Commented Dec 19, 2020 at 14:11
-
$\begingroup$ Some random thoughts: A) How do you define a SNP in the X chromosome? I wouldn't expect to identify polymorphic positions in males, so I assume you call SNP a site in which the male individual is homozyogus for an allele different from the reference. B) In line with what @ATpoint suggests, give a look at how many reads are aligned in males and females (especially in X chromosome). C) Which organism is this? Are there any reasons to suspect that females are more different from the reference than males? $\endgroup$– Fabio MarroniCommented Jan 1, 2021 at 14:19
Add a comment
|