One unitig is unexpectedly missing after HiC phasing

[zhangrengang]

The below is a unitig graph: the unitigs A and B is allelic (confirmed by aligning to reference). After HiC phasing, unitig A is in hic.hap2.p_ctg, but unitig B is missing in hic.hap1.p_ctg.

The following is the corresponding HiC heatmap:

By the way, some switch errors are observed, but they can be corrected manually.

[chhylp123]

It is possible... how long of A and B?

[zhangrengang]

About 1M bp.

[zhangrengang]

In total, two issues of this type are observed in my assembly: one is ~1 Mbp (the above) and the other is ~400 Kbp. The 400kb is somewhat different as follows: the unitig C seems to be homozygous; it is in hap1 but not in hap2 after HiC phasing.

[chhylp123]

Thanks. Can you also find unitig B in hap2?

[zhangrengang]

No. I combine hap1 and hap2 in HiC heatmap, so it will be present in the heatmap if it is in either haps.

[chhylp123]

It is interesting. Is this possible that you can share the bin file with us for debugging?

[zhangrengang]

I am sorry that the data is not my own, so I can not share.

[chhylp123]

I see. Thanks for reporting this issue.

[chhylp123]

Could you please have a try with the current release (https://github.com/chhylp123/hifiasm/releases/tag/0.16.1)? I guess it should be able to resolve this issue.

[zhangrengang]

Thanks. I have tried and the missed unitig is present this time.

[chhylp123]

Thanks a lot!