Closed ryhui closed 3 years ago
I believe your readcounts=True
is problematic. You don't actually use read count data (available via hdf5 target format), so try setting that parameter to False
. I did some updates to the readcount mode that make it behave correctly now, which in your case produce that error.
Otherwise the parameters look good.
Side note: v0.1a4 uses a faster HMM (a rescaled version not doing the log transform) and posterior0 is now saved not log-transformed. Every new run should do that automatically.
Thanks, that solved the problem!
Ruoyun
Hi Harald,
I notice the current version of hapROH gives very different result compared to the last version I used (v0.1a4) on the same data. I checked that the hap.csv and map.csv files are identical, but posterior0.csv and roh.csv are not. v0.3 is somehow tagging many more ROH than v0.1 - I got > 3000cM ROH tracks for everybody, which is almost the entire genome. This happens in both haploid and diploid mode.
It could also be user error on my side. I followed the parameters from the notebook:
Should I change anything? Also happy to share my input and output files if needed.
Best, Ruoyun