I just ran BACTpipe over ~70 samples and for some Kraken failed due to out of memory problems. Unfortunately this means prokka didn't run either. I think we should change so that prokka runs if there is an assembly output from Shovill and if there is no Kraken output to use into prokka then we just call it Unknown unknown as species classification. What do you think?
I just ran BACTpipe over ~70 samples and for some Kraken failed due to out of memory problems. Unfortunately this means prokka didn't run either. I think we should change so that prokka runs if there is an assembly output from Shovill and if there is no Kraken output to use into prokka then we just call it Unknown unknown as species classification. What do you think?