Read depth for RNA-seq based Variant Calling

Question

[vireen105]

I would like to clarify the read depth values in VCF files. I did variant calling on mouse brain RNA-seq reads using the GATK Best Practices pipeline. I would like to get the read depths at each variant for each allele but some of the depth values seem to be too low? (around 2-20 only). Is this an expected result and is the depth representative of all the original reads? Is there another way for me to get and visualize the depth values of each allele at each site?

Answer 1

[Shred]

The sum of the allelic depths may be lower than the total number of reads mapping in a given region, as not all the reads may be assigned with enough confidence to one of the detected alleles. It's explained here https://gatk.broadinstitute.org/hc/en-us/articles/360035532252-Allele-Depth-AD-is-lower-than-expected

You could visualize a single locus via IGV, multiple loci via igv-reports [https://github.com/igvteam/igv-reports]. To obtain tabular output of per-base depths I'd use perbase [https://github.com/sstadick/perbase]