Large genome result with the ultra-long data model

[ld9866]

Dear developer:

I noticed that you recently developed the latest version of Hifisam software that can be assembled using only ONT data, but I found that its effect is not as good as we expected.

We used about 100X of ONT data (library: 100kb). The pig genome we assembled was about 2.6gb in size, but we found that the assembled genome was 4.2G and had nearly 2,000 contigs. What is the reason for this?

Best regards,
Dong

[chhylp123]

How about N50? Are both haplotypes 4.2G?

[ld9866]

The Contig N50 is about 90Mb, and the number of contigs is 2775.

The version we used the (HiFi and Ultra-long) achieved 140Mb and about 130 contigs. My original intention of using this mode was to get more complete sequences. It seems that using ONT alone is not very good.

The haplotypes are about 4.3 Gb and 3.5 Gb, respectively.

[chhylp123]

Could you please show me the command line you were using, as well as the log file? Thank you so much.

[ld9866]

code: /home/lidong/Software/hifiasm-0.25.0/hifiasm -o bm2.ont.only --ont -t 256 BM2.final.combine.fastq.gz

bm2.ont.log

[ld9866]

I uploaded the code and log files. Can you see what's going on?

[chhylp123]

Got it. Thanks! Are you using R10 data with SUP base calling?

[ld9866]

Yes, the data type is what you think it is.