Best pratices for variant calling on deep sequencing TruSeq Custom Amplicon data

Question

[adp7]

Hi，

I'm trying to analyze a panel of genes in tumor samples using truseq custom amplicon (TSCA) with average depth ~2000X. I was wondering if classic tools like GATK HC, UG, MUTECT are adapted to this kind of data? Is there any extra filtering step to perform because of the PCR amplification? And if is there variant calling settings more suited for deep sequencing approach?

Thanks in advance

Answer 1

[igor]

Classic tools should be fine. However, you have to skip duplicate removal, since these are amplicons, so most reads are expected to be duplicates.

[fanli.gcb]

Would you consider this to also be the case for exome sequencing?

[igor]

Depends on the library type. Most exomes are capture-based, so the fragments should be randomly selected. There are some exceptions. For example, Agilent SureSelect is amplicon-based, so it's more like TSCA.

GATK Best Practices has the same section for WGS and WES, since the two are usually so similar.