Tetraploid genome assembly

[wyl1219]

HelIo, I used hifiasm -o rz.asm -t 30 --n-hap 4 --ul ONT.fastq.gz hifi.fastq.gz to assemble a tetraploid genome, didn't Distinguish haplotypes, would you tell me there will be 2 or 4 p_ctg.gfa in the result files and give me some suggestions? thanks very much.

[chhylp123]

If you have HiC reads, the latest release Hifiasm-0.19.3-r572 will give you 4 haplotypes. But the results might be not perfect right now.

[wjq1981]

If you have HiC reads, the latest release Hifiasm-0.19.3-r572 will give you 4 haplotypes. But the results might be not perfect right now.

Hi, when I was testing 0.19.3 I had a question, but I don't know if it's right? When I assemble a tetraploid, I use the '-l 0' parameter and the result produced by p_ctg.gfa should be a diploid size, and the hap1+hap2 produced after adding the hic data is also a tetraploid size. But when I set '--n-hap 4', the size of the resulting 4 hap file d is almost the same size as the p_ctg.gfa file. Shouldn't it be half the size of the p_ctg.gfa file here?

[chhylp123]

It is hard to say. Do you mean the p_ctg with -l0 has similar size to hap*p_ctg?

[wjq1981]

It is hard to say. Do you mean the p_ctg with -l0 has similar size to hap*p_ctg?

Yes.

[chhylp123]

Could you show the command lines?

[wjq1981]

Could you show the command lines?

/hdd/hifiasm-0.19.3/hifiasm -o ttt -t 128 -f 38 --n-hap 4 -l 0 --h1 hic_R1.fastq.gz --h2 hic_R2.fastq.gz hifi.organelle.fasta.gz 2> ttt.log

[chhylp123]

Please do not add -l0.

[wjq1981]

From the current result it is not possible to use -l0, but the old version (0.16.1) is to use -l0.

[chhylp123]

You should do not use -l0 and mannually set --hom-cov to the homozygous coverage. See: https://hifiasm.readthedocs.io/en/latest/faq.html#for-hi-c-integrated-assembly-why-the-assembly-size-of-both-haplotypes-are-much-larger-than-the-estimated-genome-size

[wjq1981]

Ok, thank you. Wish you well!

[XuDong919]

hi,haoyu.I have met the same problem, the genome size of haploidy is 1Gb, but the result of hifiasm is strange, the pctg genome size is 2.2Gb, the busco is C:99.5%[S:15.2%,D:84.3%],F:0.1%,M:0.4%,n:1614 , the hap1 genome size is 2.0GB , the busco is C:99.2%[S:16.7%,D:82.5%],F:0.1%,M:0.7%,n:1614 ; the hap2 genome size is 1.7Gb, the buso is C:99.5%[S:18.3%,D:81.2%],F:0.1%,M:0.4%,n:1614 ; the hap3 genome size is 1.9Gb, busco is C:99.0% [S:22.1%,D:76.9%],F:0.3%,M:0.7%,n:1614 ; the hap4 genome size is 1.7Gb, the busco is C:99.3%[S:21.9%,D:77.4%],F:0.2%,M:0.5%,n:1614 . this is my command: hifiasm.193 -o 21ct -t 60 --n-hap 4 --h1 HIC876_raw_1.fq.gz --h2 HIC876_raw_2.fq.gz 01m64144_220505_071151.reads.fq.gz 02m64173_220505_081857.reads.fq.gz 03m64180_220512_043802.reads.fq.gz 04m64180_220513_153507.reads.fq.gz. I have checked the log of hifiasm, the homozygous read coverage threshold is right. What should i do next ? Can I use a tool like purge_dups?

If you have HiC reads, the latest release Hifiasm-0.19.3-r572 will give you 4 haplotypes. But the results might be not perfect right now.

Hi, when I was testing 0.19.3 I had a question, but I don't know if it's right? When I assemble a tetraploid, I use the '-l 0' parameter and the result produced by p_ctg.gfa should be a diploid size, and the hap1+hap2 produced after adding the hic data is also a tetraploid size. But when I set '--n-hap 4', the size of the resulting 4 hap file d is almost the same size as the p_ctg.gfa file. Shouldn't it be half the size of the p_ctg.gfa file here?

[wjq1981]

hi,haoyu.I have met the same problem, the genome size of haploidy is 1Gb, but the result of hifiasm is strange, the pctg genome size is 2.2Gb, the busco is C:99.5%[S:15.2%,D:84.3%],F:0.1%,M:0.4%,n:1614 , the hap1 genome size is 2.0GB , the busco is C:99.2%[S:16.7%,D:82.5%],F:0.1%,M:0.7%,n:1614 ; the hap2 genome size is 1.7Gb, the buso is C:99.5%[S:18.3%,D:81.2%],F:0.1%,M:0.4%,n:1614 ; the hap3 genome size is 1.9Gb, busco is C:99.0% [S:22.1%,D:76.9%],F:0.3%,M:0.7%,n:1614 ; the hap4 genome size is 1.7Gb, the busco is C:99.3%[S:21.9%,D:77.4%],F:0.2%,M:0.5%,n:1614 . this is my command: hifiasm.193 -o 21ct -t 60 --n-hap 4 --h1 HIC876_raw_1.fq.gz --h2 HIC876_raw_2.fq.gz 01m64144_220505_071151.reads.fq.gz 02m64173_220505_081857.reads.fq.gz 03m64180_220512_043802.reads.fq.gz 04m64180_220513_153507.reads.fq.gz. I have checked the log of hifiasm, the homozygous read coverage threshold is right. What should i do next ? Can I use a tool like purge_dups?

If you have HiC reads, the latest release Hifiasm-0.19.3-r572 will give you 4 haplotypes. But the results might be not perfect right now.

Hi, when I was testing 0.19.3 I had a question, but I don't know if it's right? When I assemble a tetraploid, I use the '-l 0' parameter and the result produced by p_ctg.gfa should be a diploid size, and the hap1+hap2 produced after adding the hic data is also a tetraploid size. But when I set '--n-hap 4', the size of the resulting 4 hap file d is almost the same size as the p_ctg.gfa file. Shouldn't it be half the size of the p_ctg.gfa file here?

From my personal experience. If all your data are HIFIs data, then you can use the default parameters of this version 0.16.1 to get hap1+hap2, if the size of hap is double what you expect, you add the parameter ''-l 0''. Finally mount hap1+hap2 with hic data and then split out 4 sets of haplotype genomes.

[XuDong919]

Thank you very much for your reply. I have hifi and hic data, and I will try what you said. but i have a question, the way you spilt out of 4 sets of haplotype genomes is a tool like 3ddna, allhic?

[wjq1981]

Thank you very much for your reply. I have hifi and hic data, and I will try what you said. but i have a question, the way you spilt out of 4 sets of haplotype genomes is a tool like 3ddna, allhic?

3ddna.

[XuDong919]

ok,thanks!!

[wyl1219]

Hello, I used Hifiasm-0.19.3-r572, command lines: hifiasm -o kl.asm -t 30 --n-hap 4 --hom-cov 88 --ul ONT.fastq.gz --h1 R1.clean.fastq.gz --h2 R2.clean.fastq.gz kl-hifi.fastq.gz, then I got 1.31G hap1_ctg.fa, 1.28G hap2_ctg.fa, 0.92G hap3_ctg.fa, 1.06G hap4_ctg.fa. busco show that:

Then cat hap1 hap2 hap3 hap4 > hap1234, used juicer and 3D-DNA, the command lines are the same as before, but I only got 55.73M final.hic, the juicebox show that : I don't know where is wrong , should I only to use the p_utg.fa files? thanks, whish you well.

[chhylp123]

Probably you should open the issue in the 3D-DNA repo. It is not too bad based on my point of view.

[wyl1219]

ok, thanks.

[XuDong919]

hi,haoyu.I have met the same problem, the genome size of haploidy is 1Gb, but the result of hifiasm is strange, the pctg genome size is 2.2Gb, the busco is C:99.5%[S:15.2%,D:84.3%],F:0.1%,M:0.4%,n:1614 , the hap1 genome size is 2.0GB , the busco is C:99.2%[S:16.7%,D:82.5%],F:0.1%,M:0.7%,n:1614 ; the hap2 genome size is 1.7Gb, the buso is C:99.5%[S:18.3%,D:81.2%],F:0.1%,M:0.4%,n:1614 ; the hap3 genome size is 1.9Gb, busco is C:99.0% [S:22.1%,D:76.9%],F:0.3%,M:0.7%,n:1614 ; the hap4 genome size is 1.7Gb, the busco is C:99.3%[S:21.9%,D:77.4%],F:0.2%,M:0.5%,n:1614 . this is my command: hifiasm.193 -o 21ct -t 60 --n-hap 4 --h1 HIC876_raw_1.fq.gz --h2 HIC876_raw_2.fq.gz 01m64144_220505_071151.reads.fq.gz 02m64173_220505_081857.reads.fq.gz 03m64180_220512_043802.reads.fq.gz 04m64180_220513_153507.reads.fq.gz. I have checked the log of hifiasm, the homozygous read coverage threshold is right. What should i do next ? Can I use a tool like purge_dups?

If you have HiC reads, the latest release Hifiasm-0.19.3-r572 will give you 4 haplotypes. But the results might be not perfect right now.

Hi, when I was testing 0.19.3 I had a question, but I don't know if it's right? When I assemble a tetraploid, I use the '-l 0' parameter and the result produced by p_ctg.gfa should be a diploid size, and the hap1+hap2 produced after adding the hic data is also a tetraploid size. But when I set '--n-hap 4', the size of the resulting 4 hap file d is almost the same size as the p_ctg.gfa file. Shouldn't it be half the size of the p_ctg.gfa file here?

From my personal experience. If all your data are HIFIs data, then you can use the default parameters of this version 0.16.1 to get hap1+hap2, if the size of hap is double what you expect, you add the parameter ''-l 0''. Finally mount hap1+hap2 with hic data and then split out 4 sets of haplotype genomes.

hello，wyl1219, i have a question, you let me add the parameter ''-l 0''.When l0 is used, hifiasm produces pctg and actg. Do you combine pctg and actg and then run 3ddna?