Entering edit mode
8.7 years ago
Chirag Nepal
★
2.4k
Hi there,
Some of the annotated non synonymous SNPs (from annovar or polyphen), fall on both intron/CDS. This is possible due to alternative splicing of genes, where a part of CDS could be intron/UTR of alternative transcript of the same gene.
So, the question is, is it good practice to include such SNVs (and genes) in downstream analysis. Or just exclude them.
Thanks!
You may have to take into account the abundance of different isoforms in your tissue of interest. Depending on which isoform has been predominantly expressed, you may have to prioritize different effects of the same variant on different isoforms. For human, you may use GTEx data to find this information. Frankly speaking, this could get little trickier as you may expect different isoforms to be expressed in equal proportion.
I do not have RNA-seq data to look which isoform is more expressed. Not sure, if I should include or exclude such SNVs in the final list of high confidence SNVs. However, I see that when using GenomeMusic, it might exclude such cases, because it reported N number of SNVs were intronic.