zengxiaofei
commented
3 months ago In short, you can try higher --max_inflation values or visualize the results using juicebox or haphic plot after scaffolding in quick view mode, as discussed in FAQs:
How can I do when I see "It seems that some chromosomes were grouped together" in the clustering step?
The question is complicated. HapHiC recommends a "best" inflation parameter based on the
nchrsyou specified and the distribution of group lengths. Several factors could cause this problem.(1) The
--max_inflationparameter is too low. Try increasing it. The default value of--max_inflation 3.0is generally enough. However, sometimes the best inflations are even greater than 7.0 due to a higher background of Hi-C links between chromosomes (We encountered this in the taro genome). This could be due to biological specificity or low quality of the Hi-C library. In rare cases, if a higher--max_inflationstill doesn't work, try using quick view and manually splitting chromosomes in Juicebox.(2) Some homologous chromosomes may be grouped together due to assembly errors. This is common in scaffolding phased assemblies. Please use more aggressive parameters to correct contigs, filter out contigs, or remove Hi-C links between homologous chromosomes. If you are unsure about the type or the proportion of assembly errors, a quick view may be helpful.
(3) When the chromosome lengths vary greatly, HapHiC may mistake a large chromosome for two or more chromosomes clustered together. In this case, you can choose a reasonable inflation and run the remaining steps individually.
Other concerns:
2024-07-29 17:58:13
[stat_fragments] bin_size is calculated to be 1416101 bp 2024-07-29 17:58:15 [parse_alignments] Parsing input alignments... 2024-07-29 17:58:35 [output_pickle] Writing HT_link_dict to HT_links.pkl...
The BAM file was traversed by HapHiC within only 20 seconds, indicating that there could not be enough Hi-C links after filtering.
I used hap.p_ctg(cat hap1.bp.p_ctg.fa hap2.bp.p_ctg.fa >hap1_hap2.fa) .And I just used bp.p_ctg.fa as contig.fa and the same error was reported that 9 clusters could not be obtained.The same erro outs for using hifiasm + Hi-C to obtain hap.p_ctg or p_ctg as asm.fa.
When you used other assemblies for HapHiC scaffolding, did you realigned Hi-C reads to these assemblies?
sip123a
Hello, Dear Developer,
Thank you your effort on developing the HapHiC. I think the problem started in #15 1.The karyotype of the species： 2n=2x=18，plant. 2.My commands are as follows:
This is a diploid genome with a low heterozygosity rate, around 0.5%.The XX.hap1_hap2.fa come from:
3.And my 01.cluster always erro.When running, the following log is displayed:
The HiC.filtered.bam:
4.The methods (commands) used for Hi-C read mapping and filtering.
5.The method used for genome assembly (e.g., hifiasm + Hi-C) and the assembly utilized for scaffolding (e.g., p_ctg, hap.p_ctg or p_utg): I used hap.p_ctg(cat hap1.bp.p_ctg.fa hap2.bp.p_ctg.fa >hap1_hap2.fa) .And I just used bp.p_ctg.fa as contig.fa and the same error was reported that 9 clusters could not be obtained.The same erro outs for using hifiasm + Hi-C to obtain hap*.p_ctg or p_ctg as asm.fa.
6.Statistics for the assembly input into HapHiC
I tried to use hap1 or p_ctg.fa to run haphic pipeline defulat，But in the first step 01.cluster reported an error that the number of clustering groups was less than the number of nchr. The hifiasm +hic log file and fa_detail file are as follows: HapHiC_cluster_hifiasm(hic)_pctg.log asm.hic.p_ctg_fa_detail.log
Then,I After I add the following parameters, the output 02.reassign interrupts the operation. The log file is as follows： HapHiC_cluster.log HapHiC_reassign.log
How should I adjust the parameters? Thank you again.