Open dvg-p4 opened 1 year ago
I've tried setting the arguments as in this PR for a dockerized 1.1.6.1, and it seems to solve the issues I was seeing. I haven't explicitly done a test of this code yet--was having some difficultly installing from source locally. Is there a standard suite of tests you use for new SAIGE releases?
This solves #61, making the default max MAF set the same from the command line as it is in the R script (and claims to be in the help message).
This also acts as a workaround for #60, since if the threshold for
cateVarRatioMinMACVecExclude
MACCutoff_to_CollapseUltraRare
is consistently 10.5 instead of 10, a region will never be exactly on the cutoff. However the odd behavior would probably still occur if a user manually set those flags to integer values.I did not change
dosage_zerod_MAC_cutoff
since I have not had a chance to test with imputed sets.