Using Machine Learning for drug discovery has a huge potential to reduce cost and time to discover new drugs. In the talk a short and simplified introduction into the drug discovery pipeline is given and discussed on a specific case study for cancer. The main part of the talk consists of a detailed analysis of an approach called EquiBind to predict the 3D protein-ligand complex. All information are based on the works of Stärk et al., Ganea et al. and Lanman et al.

• https://doi.org/10.1021/acs.jmedchem.9b01180
• https://arxiv.org/abs/2202.05146
• https://arxiv.org/abs/2111.07786