This critique should be based on two papers listed below that we discussed in class on learning networks from gene expression or protein array data. Your critique should be no more than 1,500 words. Please submit your critique via Canvas. The papers are:
- Protein Interface Prediction using Graph Convolutional Networks
- Recipe for a General, Powerful, Scalable Graph Transformer
Focus only on the graph convolutional network from paper 1 that includes edge features, as described in Equation 2.
Your critique should cover each of the following points. Please dedicate at least one paragraph/section to each of these points.
1. What is the general problem area these methods address? Provide a single overview that spans the overall problem, not a separate introduction for each method.
2. Briefly describe each method.
3. What novel results and insights were obtained using these methods?
4. What are the relative strengths and weaknesses of each method?