Deep Learning for Protein Structure Prediction

Explore the cutting-edge field of protein structure prediction through deep learning in this conference talk by Iaroslav Geraskin at Conf42 ML 2024. Delve into the fundamental concepts of proteins and their importance before examining various structure prediction methods, including molecular dynamics and homology modeling. Gain insights into the evolution of AlphaFold, from its initial version to AlphaFold 3, and understand how deep learning techniques are revolutionizing this critical area of computational biology. Learn about the challenges faced in protein structure prediction, such as the three-body problem, and discover how advanced AI models are overcoming these obstacles. Conclude with a comprehensive overview of the current state and future prospects of deep learning applications in protein structure prediction.

intro
preamble
who am i
agenda
proteins
why do we need proteins?
msa
structure prediction methods
three-body problem
molecular dynamics
homology modeling
alphafold 1
alphafold 2
single seq
alphafold 3
conclusion
questions?