Scaffolding a plant genome 3d-DNA and JuiceBox

[Krešimir Križanović]

Hello,

I've been working on a genome assembly for a diploid plant genome of about 4Gbp. I've generated an initial assembly using PacBio hifi reads (25X) and hifiasm. The result was about 3100 contigs. Then I ran 3d-dna pipeline using Omni-c HiC reads, about 100X. Resulting assembly.FINAL.fasta file had over 9000 scafolds, but N50 and largest contig were improved compared to initial assembly.

I've loaded the final assembly and heat map into JuiceBox, images are attached. The heatmap looks ok to me, I just need to work manually on finishing the assembly.

I have several questions though.

- why does FINAL.fasta file have many more scaffolds that the initial assembly

- my JuiceBox is running very slow, even slowing down the server significantly, is that due to a large number of scaffolds in the final assembly

- from the zoomed image it seems that some chromosomes are made up of many small scaffolds, is there any way to merge them

- can I move chromosomes around instead of scaffolds, that would be much more convenient

- Also, where can I check the number of generated chromosomes, I'm not sure where to find that number

I actually really like JuiceBox, nice tool.

Best regards,

Krešimir Križanović

[Olga Dudchenko]

Hello Krešimir,

I recommend that you take a look at the Genome Assembly Cookbook, hosted on dnazoo.org/methods. This will help hopefully with the interpretation of the results. Your genome assembly does not look good (if that is indeed the assembly contact map rather than the draft contact map). A good assembly map looks like a handful of large squares along the diagonal. Make sure your 3d-dna run does not report problems, and if not examine why the defaults might not be working for you. See multiple threads on this forum for the description of cases when tweaking of defaults might be recommended and how they should be tweaked if necessary.

Best,

Olga