Resolving switching error (?)

[s-kyungyong]

Hi!

I am assembling a hetrozygous diploid plant genome with 110X HiFi reads and Hi-C data. The hetrozygosity is about 1%, and the genome size is 1 Gb per haplotype. After scaffolding and comparison to the reference genome, I found some sudden changes in homology between two corresponding chromosomes. In the graph below, the x-axis is my genome, and the y-axis is the reference, the green lines indicate each contig. The red box indicates such a sudden shift in homology that happened within a single contig.

After mapping the HiFi reads and investigating the regions around such breaks, I always identify a region in which haplotypes don't seem to be fully resolved as below. What would be the best way to resolve this? Some scaffolds are already good as they are now. But some others have too many to fix manually.

I used v0.19.9-r616 with the following command:
hifiasm --telo-m GGGTAAA --dual-scaf -l3 -t 56 --h1 .. --h2 .. ...

I am playing with -u and --b-cov, but in my first attempt, N50 dropped from 15 Mb to 3.2 Mb. I am not sure if this would be helpful.

Thank you!