Human Exome Variant Reference

Question

[qhtjrmin]

Hi,

I want to compare the variants for my WES analysis result using Illumina/hap.py. However I cant find the reference variants for the whole exome. I know that files (vcf, bed) in GiaB are usually used as reference variants, but I don't know which file can be used as reference variants for WES. I only could find the WGS variant references in GiaB...

I know that I can use the same reference vcf as WGS, but only change the bed file to high-confident WES parts. Where could I find the bed file for WES in GiaB?

I would appreciate it if someone let me know if the analysis method I know is wrong.

Thanks.

Answer 1

[prasundutta87]

I think you can go ahead and take up high confidence WGS variants itself as they will contain high confidence exonic variants as well. Of course, you may need to filter off non-exonic variants, as you mentioned.

Answer 2

[qhtjrmin]

Thanks for the reply, I created and used a new bed file with the intersection of exonic parts bed and high-confident WGS variants bed. However, despite using the same variant calling tool (of course, I used the version for WES), the precision, recall, and F1-score dropped significantly compared to WGS. I think there is a problem with my comparison method because the tools I used are already proven tools. Is there anyone who has an idea about this point?

Answer 3

[Jeremy Leipzig]

We offer the high-confidence regions as an option in our variant benchmarking workflow: https://truwl.com/workflows/library/Variant%20Benchmarking/0.2