How to improve the QV of the hifiasm assembly? about hifiasm HOT 3 OPEN

Yes, I guess so.

from hifiasm.

As there are higher coverage of ONT reads, the lower QV values of new hybrid assemblies might be caused by some ONT-only regions. Could you please have a try to polish these new assemblies and see if completeness could be improved?

from hifiasm.

Thanks a lot for replying me so fast! I found some error in my completeness calculation. So the 70% completeness is due to the wrong file used for data base building, and now everything has been corrected.

So most of the assemblies resulted from hifiasm have 94-98% completeness and QV around 40-53 calculated by merqury. After one run of hifi polishing by racon, the completeness only increased max 0.1, and some assemblies even have a bit lower completeness. More or less the same situation for QV. For BUSCO values, most of the complete value are above 99%, except two assemblies which have complete values around 92 or 97%. The polishing step doesn't improve the BUSCO value either.

I checked the coverage of those assemblies with relatively low BUSCO, and noticed that HiFi coverage (16x and 22x) is pretty low but Nanopore coverage is ok (31x and 60x). So I guess probably the very low coverage of hifi data for these two assemblies is the reason for low completeness?

from hifiasm.