R/README.md

In Sage-Bionetworks/fendR: Feature Engineering by Network Domain

R utilities for drug response prediction

We have experimented with a number of approaches to predict drug response from related data.

Omic data weighted with gene target data

This approach adapted from the Guan lab uses a network to weight nodes by a combination of basal expression values of the corresponding gene together with the proximity of that gene to known targets of a drug. The approach then builds a statistical model with these augmented gene features.

Development of this project can be found in the N3 directory.

Network reduction using physical drug-target-protein networks

As an alternate approach, we are experimenting with an approach that uses gene lists of interest derived from differential expression and carries out the following analysis: 1. Identify transcription factors that give rise to the gene expression changes 2. Maps transcription factor activity to protein-protein interaction network 3. Merge protein interaction network with drug target network 4. Identify proteins-drug interactions that best interrupt the active proteins

Sage-Bionetworks/fendR documentation built on May 3, 2019, 8:34 p.m.