The presentation will cover various artificial neural network methods in the field of medicinal chemistry, including tasks of molecular representation, molecular generation, properties, and conformation prediction. We will discuss different molecular representations and relevant neural network approaches, look at the most important neural architectures over the past years, and examine SOTA solutions.

Artur Kadurin graduated from Kuban State University, Department of Computer Science and Applied Mathematics year 2008. 2013-2017 he was working at the Mail.Ru advertisement department as a lead data analyst. 2017-2021 he was working on the deep learning engine for drug discovery at Insilico Medicine as a chief AI officer. In 2019 co-authored the first DL book in Russian "Deep Learning". Since 2021 he is working in Artificial Intelligence Research Institute.