DrugRPE: random projection ensemble approach

to drug-target interaction prediction



Abstract:

  Drug-target interaction playes key role in drug discovery. Since the determination of drug-target interactions is costly and time-consuming by in vitro experiments, computational method is commonly used to determine the interactions. To address the issue, a random projection ensemble approach is proposed. First, drug-compounds are encoded with feature descriptors by the use of software "PaDEL-Descriptor". Second, target proteins are encoded with physiochemical properties of amino acids, where totally 34 relatively independent physiochemical properties are extracted from 544 properties in AAindex1 database. Then the vectors of drug-target pairs are obtained. Since the vectors have different dimensions in terms of different lengths of protein sequences, random projection is adopted to project the original space onto a reduced one and thus yield a transformed vector with fixed dimension. Several random projections build an ensemble system with REPTree algorithm. Experimental results show that our method significantly outperforms and runs faster than the state-of-the-art drug-target predictors, on the commonly used drug-target benchmark sets.

 

Software available:

 A simple Matlab implement of our predictor is available here: DrugPRE.

 

Datasets:

drug-target datasets: enzymes, icon channels, GPCRs, and nuclear receptors.

 

Citation:
Jun Zhang, Muchun Zhu, Bing Wang, Peng Chen, DrugRPE: random projection ensemble approach to drug-target interaction prediction, Neurocomputing, Accepted.

Copyright @ 2004-2016 by Peng Chen

All Rights Reserved