Description

Our protein-protein interaction prediction approach is based on the training of a generative adversarial network using Wasserstein distance-based loss improved with gradient penalty performing image-to-image translation conditioned on embedding information of the network topology [1–3]. This model consists of a generator and a discriminator trained by an adversarial learning process [4].
The input for the generator is the embedding of the 90% splitted versions of the original network generated by the node2vec algorithm [5] along with a set of adjacency matrices of induced subgraphs together covering most of the nodes and edges of the original network.

Team information

Name: Olivér Balogh, MSc
Affiliation:
- Cardiometabolic and MTA-SE System Pharmacology Research Group, Department of Pharmacology and Pharmacotherapy, Semmelweis University, Budapest, Hungary
- Faculty of Information Technology and Bionics, Pázmány Péter Catholic University, Budapest, Hungary
Contribution: Write the code; Analyze the results

Name: Bettina Benczik, MSc
Affiliation:
- Cardiometabolic and MTA-SE System Pharmacology Research Group, Department of Pharmacology and Pharmacotherapy, Semmelweis University, Budapest, Hungary
- Pharmahungary Group, Szeged, Hungary
Contribution: Write the code; Analyze the results

Name: Mátyás Pétervári, PharmD
Affiliation:
- Cardiometabolic and MTA-SE System Pharmacology Research Group, Department of Pharmacology and Pharmacotherapy, Semmelweis University, Budapest, Hungary
- Sanovigado Kft., Budapest, Hungary
Contribution: Write the code; Analyze the results

Name: Bence Ágg, MD, PhD
Affiliation:
- Cardiometabolic and MTA-SE System Pharmacology Research Group, Department of Pharmacology and Pharmacotherapy, Semmelweis University, Budapest, Hungary
- Heart and Vascular Center, Semmelweis University, Budapest, Hungary
- Pharmahungary Group, Szeged, Hungary
Contribution: Write the code; Analyze the results; Project coordination

Name: Péter Ferdinandy, MD, PhD, MBA
Affiliation:
- Cardiometabolic and MTA-SE System Pharmacology Research Group, Department of Pharmacology and Pharmacotherapy, Semmelweis University, Budapest, Hungary
- Pharmahungary Group, Szeged, Hungary
Contribution: Scientific advisor, Institutional background and financing

Reference

1. Arjovsky M, Chintala S, Bottou L. (2017) Wasserstein GAN. arXiv, .
2. Gulrajani I, Ahmed F, Arjovsky M, Dumoulin V, Courville A. (2017) Improved Training of Wasserstein GANs. Adv Neural Inf Process Syst, 2017-December: 5768–5778.
3. Isola P, Zhu J-Y, Zhou T, Efros AA. (2016) Image-to-Image Translation with Conditional Adversarial Networks. Proc - 30th IEEE Conf Comput Vis Pattern Recognition, CVPR 2017, 2017-January: 5967–5976.
4. Goodfellow I, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, Courville A, Bengio Y. (2020) Generative adversarial networks. Commun ACM, 63: 139–144. doi:10.1145/3422622.
5. Grover A, Leskovec J. (2016) node2vec: Scalable Feature Learning for Networks. Proc ACM SIGKDD Int Conf Knowl Discov Data Min, 13-17-August-2016: 855–864.
6. Lee CY. (1961) An Algorithm for Path Connections and Its Applications. IRE Trans Electron Comput, EC-10: 346–365. doi:10.1109/TEC.1961.5219222.