No output of GFAse

[YouxinZhao]

The phased.fasta was from shasta13 with herro corrected reads.
However, no output of GFAse.
~/zhaobaojun/app/gfase/GFAse/build/phase_haplotype_paths -i Assembly-Phased.gfa -p paternal.fa -m maternal.fa -k 31 -o
gfase_phase_out

fasta of P and m

log

Result

[rlorigro]

Hi @YouxinZhao, it looks like there are no paths in the GFA, which are required for GFAse to run using k-mers, because GFAse was developed when Mode 2 was the standard. Unfortunately Shasta mode 3 removes this form of output. If you wanted to try to use k-mers, you could create a contact map CSV and provide that to phase_contacts_with_montecarlo, but it might be tricky to re-interpret k-mers as contacts...

I will add a warning to the ReadMe so that this is more clear.

[YouxinZhao]

The mode of conf seems like mode2

[rlorigro]

In that case you just need to use Assembly-Detailed.gfa instead of Assembly-Phased.gfa

[YouxinZhao]

How to deal with the unphased fasta?

I want to make the best use of these contigs.

[rlorigro]

Hi Youxin, If you are getting unphased contigs with the k-mer based pipeline it means that you must have very few or zero parent-specific k-mers in those contigs. This can happen because of 2 reasons: your contigs are very short, or your sequence quality is low (doesn't match illumina k-mers). The solution here is not always simple. It likely requires improving the quality of the assembly.