Guidance on CoverM vs dRep Dereplication Outputs

[sarehaghababaee]

Hello,

I ran CoverM on a set of 445 high-quality MAGs obtained using CheckM2 from 12 metagenomic datasets. CoverM performed dereplication and clustered the MAGs into 172 clusters. In parallel, I also ran dRep separately and compared its representative genomes with those from CoverM. However, I noticed that the representative genomes selected by CoverM do not fully match those identified by dRep. Could you provide guidance on how to interpret these differences?

In addition, I observed that the outputs of CoverM do not include information about the clusters themselves. I was hoping for outputs similar to dRep’s, which provide detailed cluster information. Is there a flag or option in CoverM that enables this type of output?

Thank you in advance for your help.

[wwood]

Hi,

The dereplication part of CoverM is handled by galah, which uses a different clustering algorithm to dRep, hence the different results.

You can specify more detailed clustering outputs e.g. with --dereplication-output-cluster-definition. They should all be detailed at https://wwood.github.io/CoverM/coverm-genome.html#dereplication-genome-clustering

HTH, ben