Getting higher duplicated BUSCO for a genome assembly

Question

[u.nakandala]

Hi, I have done an assembly for a few plant species using hifiasm. Complete and duplicated BUSCO values for one of the species is too high (91.4%) and complete and single copy BUSCO is very low which is 8.6% for this species. Can it be a polyploid species? But no previous records of polyploidy have been reported for this particular species previously. So we are doubtful of this result. If it is not a polyploid, what would be the other possible reasons to get this much higher duplicated BUSCOs?

[MboiTui]

Hello,

Late reply but it might be helpful to some.

One way to check if result is real or not is to use Merqury and the kmer spectra.

Aside from this, I have also experience high duplication with the initial hifi assembly. Running purge_dups and then scaffolding with SALSA or YAHS then produced better BUSCO scores and Merqury kmer spectra.

Answer 1

[h.mon]

I don't have experience with hifiasm (and if there are settings you could tweak here), but if the species is highly polymorphic, the assembler could keep contigs from both chromosomes separated, instead of collapsing them into haplotigs.