generate hic contact map based on purged genome?

[SmnYanGe]

Hi, Thank you for creating and maintaining YaHS. I would like to ask question regarding the following:
After I assembled the genome using yahs, I purged the assembly using purge_dups. I would like to know if I can use the purged genome to generate a new hic contact map, my steps are as follows:
I followed the 'https://github.com/c-zhou/yahs#generate-hic-contact-maps' procedure，
First filter the alignments_sorted.txt generated in this step to retain what is still retained in the purged assembly.
Then run juicer_tools to generate a new hic contact heatmap with new alignments_sorted.txt and new scaffolds_final.chrom.sizes, but it runs with an error:

java.lang.ArrayIndexOutOfBoundsException: Index 174 out of bounds for length 174
at juicebox.tools.utils.original.ExpectedValueCalculation.addDistance(ExpectedValueCalculation.java:171)
at juicebox.tools.utils.original.Preprocessor$MatrixZoomDataPP.incrementCount(Preprocessor.java:1670)
at juicebox.tools.utils.original.Preprocessor$MatrixPP.incrementCount(Preprocessor.java:1511)
at juicebox.tools.utils.original.Preprocessor.writeBody(Preprocessor.java:777)
at juicebox.tools.utils.original.Preprocessor.preprocess(Preprocessor.java:419)
at juicebox.tools.clt.old.PreProcessing.run(PreProcessing.java:122)
at juicebox.tools.HiCTools.main(HiCTools.java:94)

I would like to know your opinion and advice, Thanks for your help!

[c-zhou]

Hello @SmnYanGe,

I cannot tell where the problem is from the error message. My guess is that you have a very short sequence in your scaffolds - maybe 174bp long, and this triggered a juicer_tools problem. One solution could be removing those extremely short sequences. You just need to update the scaffolds_final.chrom.sizes to exclude the corresponding lines.

Best,
Chenxi

[SmnYanGe]

Hi @c-zhou and thank you for your response!

I have identified the issue: if I use purge_dups on the assembly results generated by yahs, it changes the lengths of the scaffolds, thereby causing a mismatch in data between the two files 'alignments_sorted.txt' and 'scaffolds_final.chrom.sizes'. If I simply remove some low coverage scaffoldss without performing a purge, I can successfully generate a hic map.

This leads to a second problem. I tried to manually edit the hic map, so I used the method here 'https://github.com/c-zhou/yahs?tab=readme-ov-file#manual-curation-with-juicebox-jbat' to generate a hic map of the assembly with low coverage scaffolds removed. However, it seems to be unsupported, and I receive an error message: 'Segmentation fault (core dumped)'.

From my attempts, I believe the issue is that I cannot modify the 'scaffolds_final.agp' file. I tried deleting the contents of the second scaffold only, and encountered the same error. However, when I delete the contents of the last scaffold, it runs normally. I would like to hear your opinion on this. Thanks again!

Best wishes,
Yang

[frwjo]

Hi @c-zhou and thank you for your response!

I have identified the issue: if I use purge_dups on the assembly results generated by yahs, it changes the lengths of the scaffolds, thereby causing a mismatch in data between the two files 'alignments_sorted.txt' and 'scaffolds_final.chrom.sizes'. If I simply remove some low coverage scaffoldss without performing a purge, I can successfully generate a hic map.

This leads to a second problem. I tried to manually edit the hic map, so I used the method here 'https://github.com/c-zhou/yahs?tab=readme-ov-file#manual-curation-with-juicebox-jbat' to generate a hic map of the assembly with low coverage scaffolds removed. However, it seems to be unsupported, and I receive an error message: 'Segmentation fault (core dumped)'.

From my attempts, I believe the issue is that I cannot modify the 'scaffolds_final.agp' file. I tried deleting the contents of the second scaffold only, and encountered the same error. However, when I delete the contents of the last scaffold, it runs normally. I would like to hear your opinion on this. Thanks again!

Best wishes, Yang

Hi,I meet the same promblem.But I solve the question.Because I write down the error " echo "assembly num" " ,you can check your code at this location.if you don't know how to get the num,you can try this code:awk '{s+=$2} END{print s}' your.fasta.fai
Best wishes
Wu