search

bxlab / metaWRAP

MetaWRAP - a flexible pipeline for genome-resolved metagenomic data analysis
MIT License
400 stars 190 forks source link

something about dRep #143

Open yangfanLiu1995 opened 5 years ago

yangfanLiu1995 commented 5 years ago

I have used the dRep dereplicatecommand to choose some winning bins.Then I have an idea :Can I use the bins which were not choosed to improve the winning bins' completeness?Maybe they have different defects but they can complet each other.

So is the idea right?And how can I get this?

Look forward your reply.

Yangfan 2019.3.15

ursky commented 5 years ago

If I understand dRep correctly it will just show you which bins are the same organism. The bin_refinement module of metaWRAP already does this, and it also makes sure that no contig comes up more than once in the final bin set - something you might get if you use dRep. So I would be wary of this approach.

The time to use dRep is when you have bins coming from multiple individual assemblies, that you need to consolidate. For example, if you had three samples, you could assemble, bin, and refine each set individually and together with metawrap, resulting in 4 sets of final bins coming from 4 different assemblies. Then I would use dRep to chose the best version of each bin. However if you just had one merged assembly, then do not bother with dRep - its unlikely for any two bins to be very similar because they would bet merged at the assembly level.