Hi Sergei,

Are the multiple threads implemented in HYPHYMP only suitable for some batch files? I have been trying to use HYPHYMP on dNdSDistributionComparison.bf, and although the CPU goes up to ~2000% (so it looks like the multiple threads are working), it does not run any quicker – in fact, it runs much slower than when I use HYPHYSP. This is a problem as I have some large concatenated sequences that will take longer to run than is feasible if I am constrained to using a single processor.

How does the parallelization of HYPHYMP work? Does it require different parameters (eg α1, α2; β1, β2) that can be calculated in parallel to save time? Or would it run the different models within the batch file at the same time? Either way, in the case of dNdSDistributionComparison.bf it should speed up the analysis. Have you any idea what the problem might be?

Cheers,

Daisy