Skip to content
Wide and Deep Neural Net for predicting TB resistance from genotypic data
Python R
Branch: master
Clone or download
Cannot retrieve the latest commit at this time.
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data
README.MD
evaluate_derived_features.py
evaluation.py
feature_importance.R
feature_importance.py
heatmap.R
helpers.py
lr_features.py
models.py
presel_mlp.py
requirements.txt
time_models.py
tsne.R
tsne.py
tsne_lineage.R
validation_data
validation_data.py
visualize.R
wdnn_model.py

README.MD

Deep Learning Predicts Tuberculosis Drug Resistance Status from Whole-Genome Sequencing Data.

This repository contains code to recreate the results from Chen et. al:

Chen, M.L., Doddi, A., Royer, J., Freschi, L., Schito, M., Ezewudo, M., Kohane, I.S., Beam, A. and Farhat, M., 2018. Deep Learning Predicts Tuberculosis Drug Resistance Status from Whole-Genome Sequencing Data. bioRxiv, p.275628.

Dependencies

  • Keras version >= 2.0
  • Tensorflow version >= 1.7
  • sklearn
  • numpy
  • pandas

To run the code, clone this repository and add the folder to your python path.

  • evaluation.py will recreate the cross-validation results
  • validation.py will produce the results on the validation data
You can’t perform that action at this time.