Amplicon Somatic Mutation Caller

[Buckethead84]

Hi everyone,

I'm relatively new to bioinformatics, and I am trying to use some deep amplicon sequencing data from the MiSeq to validate somatic mutations identified from either WES or other targeted sequencing methods.

In terms of experimental design, I have 3 amplicons tiled over each of the mosaic variants in question (range from 3% - 22% AAF) and an average sequencing depth of 10,000X/amplicon using 22 cycles of initial PCR & 8 cycles of indexing with Q5 polymerase.

I'm having trouble identifying the right tools to analyze this kind of data, since I am using unpaired samples and all of the reads will have the same start/end site, preventing me from being able to collapse the reads. Any suggestions? I've combed through pre-existing threads without much luck.

[henry.wood]

I had a problem like this a while back. If your tiled amplicons can be identified and separated form their start-end positions, you can split them into separate bam files. Varscan2 has the option of mutation calling from multiple input bam files, and works nicely on PCR products. You can't use the somatic mode, if you don't have the paired data, but it picks up the mutations fine. You just need to parse the output to see how many of your mutations are called in multiple amplicons.
Hope that helps. Follow up questions would need me to dig around in old code, so might take a bit longer.