John Jumper

3D-Beacons provides a federated network to unify access to protein structure data from multiple sources, enabling comparative analysis of models.

AlphaFold trained on complexes, with cross-chain paired alignments and symmetry-aware losses, gets 70% of recent heteromeric interfaces acceptable or better (DockQ ≥ 0.23), 27 points above linker-joined AlphaFold.

Enformer, a sequence-based transformer integrating elements up to 100 kb away, lifts held-out gene expression prediction (correlation 0.81 to 0.85 versus Basenji2) and variant effect prediction.

AlphaFold, an attention network over sequence alignments and residue pairs, rivaled experimental accuracy in most CASP14 cases: median backbone r.m.s.d.95 0.96 Å versus 2.8 Å for the runner-up.

The paper trains a coarse-grained forcefield using contrastive divergence on simulation trajectories to improve protein folding accuracy while maintaining computational efficiency.