Linking drug target and pathway activation for effective therapy using multi-task learning

This repository describes the code used in the manuscript

Yang et al. Linking drug target and pathway activation for effective therapy using multi-task learning. https://www.nature.com/articles/s41598-018-25947-y

Introduction to Macau

Macau installation: https://github.com/jaak-s/macau

Macau tutorial can be found here: http://macau.readthedocs.io/en/latest/source/examples.html#

Macau technical overview paper: http://ieeexplore.ieee.org/document/8168143/

Macau technical paper (extended): https://arxiv.org/pdf/1509.04610.pdf

To understand more about MCMC Sampling: http://twiecki.github.io/blog/2015/11/10/mcmc-sampling/

Drug response prediction in 4 different settings:

Setting 1: macau_GDSC.py

Setting 2: macau_GDSC_pred_new_drug.py

Setting 3: macau_GDSC_transduction.py

Setting 4: macau_GDSC_pred_new_drug_new_cell.py

You may need a certain environment to submit your job (depending on the cluster): cluster_script_GDSC.sh

Generate the interaction matrix:

Step 1: use macau_GDSC_pred_new_drug_new_cell.py (for one tissue) or macau_GDSC_interaction_tissue_QC.py (for all tissues) to check the quality of the features. Make sure the performance is greater than 0.3.

Step 2: Generate the interaction matrix using macau_GDSC_interaction_transduction.py (for one tissue) or macau_GDSC_interaction_tissue.py (for all tissues)

Step 3: use macau_GDSC_interaction_tissue_PERMUTATION.py to do random permutation of the cell lines side features. This gives p-value for each data point of the interaction matrix.

Plot the results:

Analysis feature interaction result: GDSC_interaction_analysis_Tissue.Rmd

Scatter plot of the result using: Scatter_plots.Rmd

Analyse the drug response prediction result: Drug_response_analysis.Rmd

Simply download the results:

In folder target_progeny11: interaction matrices for 16 GDSC tissues. Since drug target is binary, make sure to ONLY CONSIDER proteins targeted by at least 2 differents drugs.

In folder target_progeny11_PERMUTATION: one file about the counts generated by the permutation in order to compute the p-value ; one file for the p-value matrix ; one file for the BHY corrected p-value.

License

Distributed under the GNU GPLv3 License. See accompanying file LICENSE.txt or copy at http://www.gnu.org/licenses/gpl-3.0.html.