Presentation Information

[8p-S203-7]Machine learning in molecular interaction prediction: challenges and outlook

Taisei Watanabe1, Syuki Otsuka1, 〇Hideo Doi1 (1.Rikkyo Univ.)

Keywords:

intermolecular interaction prediction,Machine learning,descriptor

We focus on the use of machine learning in computationally expensive molecular interaction calculations. With a particular focus on interactions between biologically relevant molecules and organic compounds, we outline the descriptors used by previous machine learning models and their predicted properties. While the nature of the descriptors has a significant impact on prediction performance, their selection is not an easy task. Therefore, this presentation will provide an overview of the various descriptors and the properties they represent with examples.