I attempted to assemble a diploid plant genome using only HiFi data and assembly with hifiasm. On output of the pipeline, the final Busco and QUAST reports show a fairly low-quality assembly (80% present, 20% duplicated, 20% missing), which was quite surprising. On further inspection, I see that after hifiasm this pipeline is passing the "processed unitigs" ({prefix}.p_utg.gfa) files downstream rather than the more collapsed primary contigs. It looks like this pipeline used to pass the primary contigs downstream rather than the unitigs, but this was changed in in subworkflows/local/assemble/main.nf in commit 71a1c85 after the hifiasm nf-core module was updated. I can think of a few cases where we would want the unitigs passed downstream, but I imagine in most cases (certainly in my own) people would still want the primary contigs passed instead. Can there be an option to change this behavior?
I attempted to assemble a diploid plant genome using only HiFi data and assembly with hifiasm. On output of the pipeline, the final Busco and QUAST reports show a fairly low-quality assembly (80% present, 20% duplicated, 20% missing), which was quite surprising. On further inspection, I see that after hifiasm this pipeline is passing the "processed unitigs" ({prefix}.p_utg.gfa) files downstream rather than the more collapsed primary contigs. It looks like this pipeline used to pass the primary contigs downstream rather than the unitigs, but this was changed in in subworkflows/local/assemble/main.nf in commit 71a1c85 after the hifiasm nf-core module was updated. I can think of a few cases where we would want the unitigs passed downstream, but I imagine in most cases (certainly in my own) people would still want the primary contigs passed instead. Can there be an option to change this behavior?