search

anuradhawick / LRBinner

LRBinner is a long-read binning tool published in WABI 2021 proceedings and AMB.
https://doi.org/10.4230/LIPIcs.WABI.2021.11
GNU General Public License v2.0
29 stars 5 forks source link

Questions about parameter optimization for strict binning #22

Closed franciscozorrilla closed 4 months ago

franciscozorrilla commented 4 months ago

Hello, thank you for developing this tool!

I ran some metagenomic nanopore-based flye assemblies through LRBinner, and then ran the resulting bins through checkm, but its looks like most bins are over 100% conaminated. Questions:

  1. I was wondering which parameters I should adjust in order to get more strict binning?
  2. Would you expect that binning reads prior to assembly might help me here?
  3. I am also seeing larger genome sizes than expected based on the bin bp content, is this a common issue perhaps related to duplicated contigs?
  4. Is there a newer/improved nanopore binning tool that you or others have developed that I should try?

This is the command i used for lrbinner contig with default params:

ls assemblies/|while read sample;do 
    python LRBinner/LRBinner/lrbinner.py contigs --reads-path fastp/${sample}.fastq -k 3 -t 8 -sep -o bins/${sample} --contigs assemblies/${sample}/assembly.fasta;
done

Thanks and best wishes, Francisco

anuradhawick commented 4 months ago

Hello. Thanks for trying this out.

The contigs binning is not a stable feature. This was initially built for long reads binning.

You might have better luck trying

https://github.com/metagentools/MetaCoAG