Pavel Kohout

Superior Biocatalyst Design using Latent Spaces of Artificial Neural Networks

Lector: Dr. Stanislav Mazurenko

Institution: Masaryk University, Faculty of Science

Field: Environmental Health Science

 

About the project

Unsupervised machine learning makes it possible to exploit large amounts of unannotated biological data. Latent space-based models can map input proteins onto an internal coordinate space to reveal their meaningful representation. This information can be used to guide rational protein design by suggesting new protein sequences or identifying important residues for mutation.

 

Why science?

Science is a way to satisfy my curiosity, to work on interesting and diverse projects, and to have the space to develop myself and new algorithms on problems that the industry doesn't usually have time for. All this with the idea that my work can become part of the solution to global problems and have a positive impact on our society.

 

What do I like most about Brno

I was born, grew up and studied in Brno, and I have a lot of very good friends with whom I have shared countless wonderful experiences. In addition, Brno, with its size, quality institutions and beautiful nature, is a place that offers everything necessary for a happy life.