Andriy Kryshtafovych

The paper describes a Kaggle competition for RNA 3D structure prediction, showing that template-based methods can outperform deep learning approaches.

This paper evaluates progress in predicting protein complex structures in CASP16, finding moderate improvements but ongoing challenges in model ranking, stoichiometry prediction, and handling difficult targets.

Across 42 CASP16 nucleic acid targets, accuracy stayed tied to template availability; human experts (Vfold first) led servers and AlphaFold 3, though servers now improve on templates on average.

This paper evaluates methods for modeling multi-state protein and RNA systems in CASP16, finding improved accuracy for some targets but persistent challenges in capturing structural details for others.

Structure providers for 18 CASP16 nucleic acid targets find blind predictions capture secondary structure but miss non-canonical interactions, backbone kinks, ligand sites, and intermolecular interfaces.

This paper assesses CASP16 monomer structure prediction, showing near-solution for single-domain folds but highlighting challenges in truncated sequences and model ranking.

The paper evaluates CASP15 cryoEM predictions using density maps, showing high accuracy but challenges in side-chain positioning and loop modeling.

The paper evaluates AI-based protein assembly prediction in CASP15, showing improved performance over CASP14, with community models outperforming the baseline AF2-Multimer in many cases.

The paper uses petabase-scale homology searches from SRA data to improve protein structure prediction accuracy using ColabFold, achieving better results than default methods.