search

ruanjue / wtdbg2

Redbean: A fuzzy Bruijn graph approach to long noisy reads assembly
GNU General Public License v3.0
513 stars 94 forks source link

Splitting FIXP kmers, assertion issues and parameter queries #130

Closed harish0201 closed 5 years ago

harish0201 commented 5 years ago

Hi!

I'm still assembling the genome on a parameter sweep as per our last discussion.

I'm getting this error, for the parameter: -g 2.5g -X 60 -p 17 -k 6 -L 5000 -AS 4 -o run1 --aln-dovetail -1

I used the dovetail metric from another thread on the issues a while back and wanted to try this out.

0|0wtdbg2: kbm.h:511: split_FIXP_kmers_kbm: Assertionrs[0]->n_head > 0' failed.`

Is there a particular reason that this happens?

The genome is around 2.5g with an estimated 80% repeat content and about 2% heterozygosity.

I have over 190Gb data in reads over 5Kb in length on Sequel. I did try the default -x sq set and the assembly is just a bit less than 2Gb which the reason I'm doing the sweep.

This is the attached kmer histogram on defaults. Can you suggest some parameters to tweak?

The parameters passed are: -p 21 -AS 2 -g 2.5g -L 5000

image

ruanjue commented 5 years ago

I noted that kbm.h:511 in latest commit is not this assert, please update your progam. If it occurs again, I will debug it.

harish0201 commented 5 years ago

Sure, thanks for the help, do you have any particular commit that I have to use?

ruanjue commented 5 years ago

Try add -R, otherwise I have no idea.

harish0201 commented 5 years ago

Ok. I'll try again with the latest release.