multiomics-t1d-ml

• stacking.py for proposed multi-view ensemble model, as shown in paper.
• supervised.py testbed for supervised LOOCV experiments. Please see comments to toggle between experiment for real data and generation of null distribution.
• generate_null.py for collecting results from feature importance experiment (relies on supervised.py)
• supervised_sig.py extraction of consistent feature sets

data directory contains source parallel multi-omic T1D datasets. The directory also contains experimental results from one run of the feature importance experiment.