Test-drive the GATK tools and Best Practices pipelines on Terra
Check out this blog post to learn how you can get started with GATK and try out the pipelines in preconfigured workspaces (with a user-friendly interface!) without having to install anything.
Identifying Rare SNPs
I have been reading through GATK's docs and forums trying to wrap my head around the best way to approach the problem of sequencing viral populations. In this case, you have sequenced a sample comprised of an unknown number of genetic variants, and want to accurately identify point mutants. For example, a mutation present in a very low percent of the genomes sampled (ie. reads) could be real. If you sequence a diploid organism, you have only 2 possibilities at each position. We have no idea how many possibilities exist at each position of the viral genome. I have found this to be a difficult problem for most SNP callers, or at least I have not found the right sort of configuration to produce descent results. I have been trying to find parallels in other areas, and have been trying to find right keywords to use in searches, but have not had a lot of luck.
It occurred to me that there may be a parallel for some cancer studies. If you take a patient sample, some cells may be normal and some cells may carry the somatic mutation. It is a heterogeneous population, just like a viral swarm. The degree of heterogeneity will obviously depend on the nature of that sample. This may be a shot in the dark.
Does anyone have experience with this sort of problem, or do you have suggestions on areas or keywords where I should be looking? Note: I did find VPhaser2, which does look interesting.
Thanks in advance for any help or pointers.