search

RolandFaure / Hairsplitter

Software that separates very close sequences that have been collapsed during assembly. Uses only long reads.
GNU General Public License v3.0
26 stars 0 forks source link

raven fails #8

Closed kazumaxneo closed 6 months ago

kazumaxneo commented 6 months ago

Hi.

This is a great implementation. When running Hairsplitter, however, it fails on the raven reassemblies after a while (attached picture).

$python hairsplitter.py -f SRR27955205.fastq -i metaflye_assembly.fasta -x ont -o outdir -t 20

I am not sure what the cause is, so I would appreciate it if you could include tiny dataset in the repository or write an example of some public data run if possible.

Kazuma Uesaka

screen-shot

RolandFaure commented 6 months ago

Hello Kazuma,

I'm very glad to see you like GenomeTailor. If I understand well this issue is in the wrong repository and concerns GenomeTailor alone ? (the GenomeTailor version packed with HairSplitter should not have these problems)

Thank you for bringing up the issue with raven, it was a compatibility issue with different versions of raven that should be corrected now.

The repository of GenomeTailor was a bit messy since I did not realize anyone actually used it except me, sorry for that. I cleaned it up, now it should work better. You were absolutely right about the test instance, I added it.

Keep in mind that GenomeTailor is in beta version , that the reassembly module does not work very well for now and that I am actively developing it. I hope to have a significantly improved version by the end of next week. Nevertheless, all you comments are welcome to help improve GenomeTailor :smiley:, if you desire a particular feature it is the best time to say so!

kazumaxneo commented 6 months ago

Hi RolandFaure

Thank you ! I had a little concerned and cloned GenomeTailor repo. independently. That would caused this problem. When I used GenomeTailor enclosed with HairSplitter as you commented, the problem did not happen (job finished successfully).

I noticed one another thing : "samtools consensus" commnad looks introduced from samtools v1.16. So I think it would be better to use samtools > v1.16 in the install section of README (currently 1.19 is the latest).

Take care.

RolandFaure commented 6 months ago

Hello Kazuma, Glad to hear the problem is solved. You are absolutely right for the samtools version, thank you very much for bringing it up :smile: Do not hesitate if you have more questions, problems or suggestions