psipred / STRING2GO

Protein function prediction based on protein-protein interaction network topology and deep maxout neural networks

STRING2GO

This is a Keras implementation of STRING2GO method reported in a published paper:

Wan, C. Cozzetto, D. Fa, R. and Jones, D.T. (2019) Using Deep Maxout Neural Networks to Improve the Accuracy of Function Prediction from Protein Interaction Networks. PLoS One, 14(7): e0209958.

---

Requirements

• Python 3.6
• Numpy
• Keras (Theano backend)
• Scikit-learn

---

Running

• Step 1. Generating network-embeddings of STRING network using Mashup [1] or node2vec [2] methods. The generated embeddings can be found in the ./data folder.

  • [1] Cho et al., (2016) Compact Integration of Multi-Network Topology for Functional Analysis of Genes, Cell Systems, 3 , 540–548.
  • [2] Grover A. and Leskovec, J., (2016) node2vec: Scalable Feature Learning for Networks, ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD).

• Step 2. Learning functional representations using ./src/STRING2GO_Functional_Representation_Learning.py.

• Step 3. Training support vector machine library for predicting protein function using ./src/STRING2GO_Functional_Representation_SVM.py.