MetaBAT 2 output varies from one run to another

[Former user]

I run MetaBAT 2 on the same dataset multiple times, and the number of bins, as well as the size of the bins, varies from one run to another. Is this normal? If so, how should I proceed? Thanks.

[Rob Egan]

There are a few places in the algorithm where randomization is used, you can set the --seed option to be non-zero and then the results between runs should be identical (assuming identical inputs).