Open iskandr opened 7 years ago
May also want to consider https://sites.google.com/site/bioericscript as it's what UCSC's ProTECT uses for fusion calling in their neoepitope selection pipeline.
I actually did this at some point: https://github.com/hammerlab/biokepi/commit/4f3449d4f7bd06bde609b7a759daf6d59e883ad6
I'm 80% certain this worked, but will re-test and submit a PR. The issue was that on many, many BAMs this required > 60 GB memory and would fail, which was really unfortunate.
Grab a highmem
type! https://cloud.google.com/compute/docs/machine-types
Yea that is actually what spurred my move to gcloud originally, got a 128G machine to run all of our samples through.
Identifies fusion proteins from RNAseq and also catches expression of viral sequences:
https://github.com/ndaniel/fusioncatcher