AI for Science | Saka Research

[TODO: review this text.]

Many systems in the sciences are naturally relational and geometric, from molecules to materials and cells. Together with our colleagues at AITHYRA, we develop and apply machine learning methods to these domains, with a particular focus on the life sciences.

Key Directions

Molecules and drug discovery: Generating macrocycles and other molecules, and graph learning over biomedical knowledge graphs for drug discovery.
Materials and engineering: Geometric deep learning for materials and structural steel design.
Dynamical systems: Learning the dynamics of natural processes, such as cell dynamics, from population-level observations.

ai-for-sciencemoleculesmaterialslife-sciences