Drug-Drug Interaction Prediction by Neural Network Using Integrated Similarity

Due to the great importance of this issue in the economy, industry, and health, proposing appropriate computational methods for predicting unknown DDI with high precision is challenging. We propose a novel machine learning method for predicting unknown DDIs called "NDD", using a two-layer fully connected neural network. NDD uses various characteristics of drugs to have comprehensive information. Multiple drug similarities are calculated. NDD integrat various drug similarities with a non-linear similarity fusion method called "SNF" to achieve high-level features.

Link of paper: https://www.nature.com/articles/s41598-019-50121-3

Dependency:

• python version 3.5.3
• keras lib https://github.com/fchollet/keras/
• scikit-learn

Codes and Data

Find DS1-Ds3 on NDD folder. Codes of functions of NDD are available in NDD folder.

Contact

Please do not hesitate to contact me if you have any question:

Email: [email protected]

Please cite us if you find this study helpful.