A Biomedical Knowledge Graph based Method for Drug-drug Interactions Prediction through Combining Local and Global Features with Deep Neural Networks
A deep learning-based DDIs predictor
This paper includes four datasets that DS1, DS2, D-DS3 and E-DS3 processed from reference [1-3] . Each dataset file contains the three feature files and the results of proposed model. Furthermore, There are examples of saving models each fold in the path of " 'dataset name' / 'method.h5' ".
python==3.8.5 tensorflow-gpu==2.2.0 pandas==1.2.4 numpy==1.20.2 keras==2.4.3 gensim==4.0.1 tqdm==4.63.0 h5py==2.10.0
We provided the example to run our code in the fold of " 'example' / 'test.py' ". You should move the feature files into 'example' / 'representation/' " and run the " 'test.py' ". If there is any questions or needs, please contact me by leaving message.
Zhong-Hao Ren, Zhu-Hong You, Chang-Qing Yu, Li-Ping Li, Yong-Jian Guan, Lu-Xiang Guo, Jie Pan, A Biomedical Knowledge Graph based Method for Drug-drug Interactions Prediction through Combining Local and Global Features with Deep Neural Networks
A Biomedical Knowledge Graph based Method for Drug-drug Interactions Prediction through Combining Local and Global Features with Deep Neural Networks
[1] Yue, X.; Wang, Z.; Huang, J.; Parthasarathy, S.; Moosavinasab, S.; Huang, Y.; Lin, S.M.; Zhang, W.; Zhang, P.; Sun, H.J.B. Graph embedding on biomedical networks: methods, applications and evaluations. 2020, 36, 1241-1251.
[2] Zhang, W.; Chen, Y.; Liu, F.; Luo, F.; Tian, G.; Li, X.J.B.b. Predicting potential drug-drug interactions by integrating chemical, biological, phenotypic and network data. 2017, 18, 1-12.
[3] Shi, J.-Y.; Mao, K.-T.; Yu, H.; Yiu, S.-M.J.J.o.c. Detecting drug communities and predicting comprehensive drug–drug interactions via balance regularized semi-nonnegative matrix factorization. 2019, 11, 1-16.