Using the MICoNE python library to infer final consensus networks

[fantin-mesny]

Hello,

I wanted to use the MICoNE pipeline (NI only), but when running micone install, 8/10 mamba installations fail due to conflicts in dependency versions.

I was wondering if I can install and run softwares outside of the pipeline, and then use the MICoNE python library to obtain consensus networks using the SV and/or SS algorithms.
If it is possible, can you please let me know how to parse data resulting from the different softwares and which python functions I should use for consensus network assembly? I could not find much documentation about the Python library. Please let me know if I missed a relevant page.

Many thanks, and best wishes,
Fantin

[dileep-kishore]

Hi Fantin,

If you want to use the consensus network algorithms, you can simply install the MiCoNE Python package by running pip install micone. Make sure your interaction data and metadata are in the correct format (https://micone.readthedocs.io/en/latest/data_import_export.html). Once that's done, you can use the Network object to load the network. To generate the consensus, you'll need to create a NetworkGroup object using a list of Network objects, as demonstrated in the example.

The scripts that are executed during the Nextflow pipeline run can be found in this folder: https://github.com/segrelab/MiCoNE/tree/master/micone/pipelines/templates/network_inference/network. You can run them manually after formatting your input data correctly.

Best,
Dileep