Almost 1/3 of the obtained bins from BASALT were removed by dRep dereplicate -sa 0.99.

[zzz199125]

In my case, although BASALT obtained much more bins (115) than metaWRAP (75), much of them are identified as the same species by GTDB. And after treated by dRep dereplicate -sa 0.99, 79 bins were finally obtained, which was as many as metawrap in number. Thus， the effectiveness of de-replicate algorithms for removing redundant genomes in BASALT is debatable, and I doubt BASALT recovers more and higher quality MAGs from benchmark data than other binning tools due to this reason!

[camillaln]

I also end up with several bins with ANI > 0.99. In my latest run BASALT gave 113 bins that were reduced to 86 after running dRep. It would be great if you improved the de-replication step.