Implementation of MTNNGC_ADMET in KNIME

Groups Cheminformatics
1

Hi,
I have a project that is actually quiet similar (similar steps, conda packages and models) to the MTNNGC_ADMET on GitHub (GitHub - fmonta/mtnngc_admet: Code for training and inference for multitask graph convolutional networks) that I want to integrate in a KNIME workflow but I am not sure how to start. Integrating the conda environment should be easily done with the “Conda Environment Propagation” node but how do I integrate the different modules that are installed with the setup.py. I am just wondering if there is a similar knime workflow that has already implemented sth, similar as in MTNNGC_ADMET (GitHub - fmonta/mtnngc_admet: Code for training and inference for multitask graph convolutional networks).

I would appreciate any advice.

Many thanks.

Best

Anjo

2

I might try sth. like this: Use functions and Python code from an external .PY file or a Juypter notebook in KNIME's Python Source node – KNIME Hub. Install the packages with the conda .yml file and the modules with “pip install -e .” and then use the python source node?