Deep Mining Heterogeneous Networks to Predict Novel Drug-Target Associations
Objective:
We propose a similarity-based drug-target prediction method that enhances existing association discovery methods by using a topology-based similarity measure.
Methods:
- NetWork: Linked Tripartite Network (LTN)
- Similarity measure: DeepWalk
- Inference method: DBSI, TBSI
Usage
- Generate deepwalk index with DeepWalkMethod.java
- Predict with Prediction.java
- For nodes that are not listed in the network (new drugs or targets), use the similarity measures in chemicstrc and genomicsqs
Examples
please check src.edu.ucsd.dbmi.drugtarget.main.Job.java for usegae
Contact
For help or questions of using the application, please contact [email protected]
Citation
The authors appreciate to cite our published work,
Zong, N., Kim, H., Ngo, V. and Harismendy, O., 2017. Deep mining heterogeneous networks of biomedical linked data to predict novel drug–target associations. Bioinformatics, p.btx160.