Too many contigs and confusing genome size

[Lillian-21]

Hi,

Thanks for you nice hifiasm!

I have a haploid plant sample (wu just guess from previous researches and experience). The WGS estimatics size is ~2.1G with the first kmer peak, the hifi data estimatics size is ~500Mb by genoscope2 with ploidy 2. The kmer frequency results are as follows.

So, it is so confusing about the genome size and ploidy. Could you give me some advise?

I have test many times with different parameters, including with "-l0" or not, with hic data or not, modify -s -D, but all the results are fragment, incomplete and high depetition. The following id my assembly results (*p_ctg.fa) and log files.

hifasm.sh.e531543.txt hifasm.sh.e1188977.txt

How can I improve my assembly? and should I use the "-l0" or not? Looking forward with your reply. Best wishes!

[chhylp123]

I guess this one is not a haploid genome. I recommend you check a few duplicate genes to see if they are real.

[Lillian-21]

Thanks for your reply.

Yes, we also doubted with the ploidy. Becuase the contigs is too fragment , I only can plot the synteny comparison with two gene sets. From the results, we can find obvious duplication.

[zxgsy520]

This is likely to be polyploid, and the three peaks have an obvious fold relationship.

[Lillian-21]

Thanks. Yes, I think so it is a polyploid (triploidy? ) now. I drawed the haplotype structure with heterozygous kmer pairs by Smudgeplots, the results show clear triploidy. And I also have a try to anchor the hifiasm's results to chromosomes with HIC reads, but the results were not ideal. It's obviously to find two parts, and I also don't know what the white area (without anchor? ) is between the two parts.

Whatever, I want to get a high quality haploid assembly results. Could all of you give me some suggestions?