Entering edit mode
5.0 years ago
lait
▴
180
Hi,
I have WES data for a sample having a specific phenotype. I ran EXCAVATOR on that sample, trying to detect CNVs. I am interested to check deletions happening on gene A, because patients having that specific phenotype would frequently get deletions in that gene A.
In EXCAVATOR's results file: FastCallResults_Sample, I found one big deletion covering all of gene A.
When I cheked this sample on IGV, I saw that the exons of gene A are well covered, with a minimum coverage of 30 reads and reaching up to 100 reads.
My question is, is this contradicting? or did I misunderstand EXCAVATOR's results ? is there an explanation for this scenario?
The gene may have coverage but how is the gene's coverage when compared to the coverage overall? Are you sure that EXCAVATOR is saying complete deletion, or is it just deletion of just one allele?
Calling CNV from NGS data is fraught with bias and error. For one, the caller has to correctly model GC bias, repeat sequence, and sequence homology - I don't know EXCAVATOR well enough to know if it manages these sources of bias or not.
The coverage of the gene is almost the same acrross the other exons.. EXCAVATOR shows a genotype of 1/1 for that deletion.. actually it shows a genotype of 1/1 for all the detected CNVs, whether it is a deletion or an insertion. Do you recommend another good SV caller for exome data to try if the results overlap somehow?
Sorry, I don't know enough about EXCAVATOR to comment further. That said, the 1/1 implies that EXCAVATOR has called virtually all indels as homozygous, meaning, on both alleles. I will ask re: SV caller