Currently working on improving the way epidisco handles all the MHC predictors and trying to make it much more fail-safe (since all the netMHC tools act in weird ways). One of the things I would like to get in there is to validate the list of MHC alleles (usually inferred via HLA typers) passed down to the predictor and make sure we don't pass anything that the tool doesn't support.
This PR adds CLI options to extract the list of supported peptides/lengths from mhcflurry and will help us during the mhcflurry's eventual integration to the pipeline. Since I felt bad making your CLI arguments a bit longer, I also started grouping them in the help text with some argparse foo.
Coverage decreased (-0.08%) to 75.378% when pulling 352efd2f2a306c5ecc0ace57ae6b307b39f9c22e on get-supported into 8b07269a0f621e8de061f197b4bf768dfb9a0cc8 on master.
Currently working on improving the way epidisco handles all the MHC predictors and trying to make it much more fail-safe (since all the netMHC tools act in weird ways). One of the things I would like to get in there is to validate the list of MHC alleles (usually inferred via HLA typers) passed down to the predictor and make sure we don't pass anything that the tool doesn't support.
This PR adds CLI options to extract the list of supported peptides/lengths from mhcflurry and will help us during the mhcflurry's eventual integration to the pipeline. Since I felt bad making your CLI arguments a bit longer, I also started grouping them in the help text with some
argparse
foo.