search

biocore / mmvec

Neural networks for microbe-metabolite interaction analysis
BSD 3-Clause "New" or "Revised" License
118 stars 51 forks source link

Exact parameters used in paper #134

Closed LeonurusCardiaca closed 4 years ago

LeonurusCardiaca commented 4 years ago

Hi!

I want to see if I can reproduce your results from the paper, therefore, I was wondering what the exact parameters are that you used for the mmvec runs for the benchmarking and ihmp study. I looked through all the files in the multiomic-occurrences github, but I couldn't find any documentation on which parameters exactly were used to obtain the best model.

Thanks in advance!

mortonjt commented 4 years ago

@isisannelie this is a little embarrassing, looks like those scripts were accidentally left out.

I've created a pull request here: https://github.com/knightlab-analyses/multiomic-cooccurrences/pull/15 We'll need first make sure that paths match up before merging.

For future reference, we'll want to have all benchmark related issues on the multiomics-cooccurrence repository so that we keep the mmvec features / bugs separate from reproducing the manuscript results.

mortonjt commented 4 years ago

Hi @isisannelie, I've merged in the missing scripts in the multiomic-co-occurrences repository. Thank you for bringing this to our attention, please feel free to raise more issues on that repository if you run into any additional problems.