Entering edit mode
3.6 years ago
DKA
▴
40
Hello, I ran BCFtools stats on a cohort of human VCFs generated from WES; however, the average TiTv (Transition/ transversion) ratio was around 2.4, does this ratio grantee that QC is good or not, please? and if not, for what does it indicate? I read that the average TiTv ratio should be around 3 for human WES; nevertheless, the average ratio was 2.4 in my case, does this mean that there might be a chance for false positives calls? Thank you
Regarding TiTv ratio being a good metric for QC and all questions regarding what value is good for which experiment, this paper may be helpful: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4308666/
If you have already removed low quality variants from your dataset, for WES, the ratio should be around 3. Since, you have a TiTv ratio of aound 2.4, it is possible that you have either not removed low quality variants or there are non-exonic variants present in your variant dataset.
WES is never accurate because you will also get variants from intronic regions or non-target regions (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3416685/), which will reduce your overall TiTv ratio to around 2.5. If you keep only exonic region variants and remove poor quality variants, you should get your TiTv ratio to be around 3 or more.
That is really helpful. Thank you so much.