reassembling chr0

[mictadlo]

Hi,
I ran RaGOO, and 2,288,581,979 bp could be mapped to the reference, but 524,739,651 could not be aligned. How would it be possible to use the unmapped contigs and reassemble them?

Thank you in advance,

Michal

[malonge]

It does seem that your chr0 is a little too large. May I ask what genome you are assembling? Also, what is your contig N50?

[mictadlo]

Hi,
The reference is an allotetraploid plant 3G genome. Please find below the statistics with different contings length filtering.

[malonge]

Am I interpreting this right by observing that you have a contig N50 of 5k? If you have a large number of very small contigs, those are unlikely to get placed by RaGOO. Only alignments over 1k are even considered. What is the size distribution of contigs that are unplaced?

Also, with these difficult plant genomes, minimap2 may not always produce good alignments. What is the distribution of quality values in the 12th column of the contigs_against_ref.paf file?

If you have the time, you can also try aligning the genome with nucmer. With nucmer, one can set very strict parameters that can sometimes help with polyploid genome alignment. I would recommend -l 200 -c 500. You can even bump -l even higher to get higher specificity in alignments. If you decide to try this, I can show you how to replace the usual minimap2 alignments with these nucmer alignments.

[dandy3745]

Hi,
I'm interested in your last reply and I'd like to get good comments from you.
We're studying hexaploid plant, and we'd like to do reference-guided scaffolding using hexaploid contigs with complete diploid plant genome.
Could you teach me how to replace minimap2 to nucmer in RaGOO? or Do you have any good ideas for this work?

Thank you,
Jung

[malonge]

Hi there!

The alignments used for scaffolding are in ragoo_output/contigs_against_ref.paf. RaGOO will not recreate that file if it already exists, so the trick is to replace that file with nucmer alignments.

First, run Nucmer to get your delta file and covert the delta file to a PAF format with paftools. Then, just rename that file to contigs_against_ref.paf and put it in ragoo_output.

Version 2, which will hopefully be ready in the next few months, will allow users to use Nucmer directly.