Entering edit mode
7.6 years ago
dana107
•
0
My overall goal is to perform CNV analysis on whole genome sequencing using a combinatorial approach (read depth + read pair or split pair). I am looking for rare events for Mendelian disease. Most of the CNV callers I have explored require population level data or paired tumor-normal samples. There are a number of callers I would like to try but are specified for paired tumor-normal samples. Is it possible to use this with the germline sample of interest as the 'tumor' sample with a sample processed in a similar way for the 'normal' control? Or perhaps use a gold standard genome, such as Illumina's platform genomes or Genome in a Bottle samples?
Did I understood correctly that you want to compare 1 sample to another sample? Best would be to have a larger cohort.
Yes, compare 1 case sample to 1 control sample. I agree, but we do not have enough WGS yet to use a population-based caller, such as GenomeStrip. As I understand, it seems most CNV callers are either for somatic CNV with an input of case-control paired samples (tumor and germline), or population based callers that use the distribution of a larger cohort to call CNVs. I did use CANVAS for calling on a single germline, but it is based on read depth only - I would like a combinatorial approach. Any suggestions?