Geometric ML becomes real in fundamental sciences

 By Medium - 2020-12-30

Description

How? A graph neural network was trained to predict the growth inhibition of the bacterium Escherichia coli on a dataset of >2000 molecules (including approved antibiotic drugs as well as natural…

Summary

  • The best of Graph Deep Learning in 2020 Geometric ML becomes real in fundamental sciences Among many papers on Geometric and Graph ML, applications in biochemistry, drug design, and structural biology shone in 2020.
  • AlphaFold 2.0 is an “attention-based neural network” (likely a transformer architecture) trained end-to-end on 170000 protein structures from the Protein Data Bank as well as protein sequences with unknown structure.
  • Since proteins are common targets for drug therapies (typical drugs are small molecules designed to bind to their target), the pharmaceutical industry has a keen interest in this field.
  • The network is trained using a few thousand co-crystal protein 3D structures from the Protein Data Bank to address multiple tasks, including interface prediction, ligand classification, and docking, showing state-of-the-art performance.

 

Topics

  1. Machine_Learning (0.44)
  2. Database (0.27)
  3. Backend (0.18)

Similar Articles

CVPR 2020 Underway, Best Papers Announced

By Synced | AI Technology & Industry Review - 2020-06-16

The 2020 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) has announced its best paper awards.

Self-Organising Textures

By Distill - 2021-02-27

Neural Cellular Automata learn to generate textures, exhibiting surprising properties.