Hahnbeom Park

G-screen: Scalable Protein-Aware Virtual Screening through Flexible Ligand Alignment

G-screen provides a scalable virtual screening method that combines flexible ligand alignment with protein-aware pharmacophore analysis to efficiently filter large chemical libraries.

ProteinSketch translates spatial intuition into protein design with bimanual interaction in VR

ProteinSketch enables interactive protein design in VR by translating spatial intuition into constraints for AI-driven protein structure generation.

ProMiSE: Protein Multi-State Evaluation Benchmark in Biological Contexts

ProMiSE evaluates protein multi-state prediction models using a curated dataset of biological conformational changes, revealing limitations in capturing functionally relevant dynamics.

Using experimental results of protein design to guide biomolecular energy-function development

The paper uses experimental results from de novo protein designs to identify and resolve failure modes in the Rosetta energy function, improving its accuracy in predicting stable protein structures.

Protein folding stability estimation with explicit consideration of unfolded states

IFUM is a deep learning model that improves folding stability prediction by jointly estimating free energy and unfolded state ensembles, showing better accuracy than previous methods.

Deep learning molecular interaction motifs from Receptor structure alone

This paper introduces MotifGen, a deep learning model that predicts molecular interaction motifs from receptor structures, enabling peptide binder design and small-molecule binding site prediction.

SHARP: Generating Synthesizable Molecules via Fragment-based Hierarchical Action-space Reinforcement Learning for Pareto Optimization

SHARP generates drug-like molecules using reinforcement learning with a fragment-based action space, improving synthesizability and binding affinity over existing methods.

Physics-inspired accuracy estimator for model-docked ligand complexes

The paper introduces DENOISer, a deep neural network that improves scoring accuracy in model docking by incorporating physical knowledge, outperforming existing tools on benchmark sets.

An artificial intelligence accelerated virtual screening platform for drug discovery

The paper develops RosettaVS, an AI-accelerated virtual screening platform that identifies drug candidates with high binding affinity across two targets, achieving a 14% and 44% hit rate in under seven days.

Accurate prediction of protein structures and interactions using a 3-track network

The paper introduces a 3-track network for accurate protein structure and interaction prediction, achieving performance close to DeepMind's CASP14 results and enabling rapid modeling of protein complexes from sequence data.

Improved protein structure refinement guided by deep learning based accuracy estimation

The paper introduces DeepAccNet, a deep learning framework that improves protein structure refinement by estimating accuracy and guiding Rosetta refinement.

Improved protein structure prediction using predicted inter-residue orientations

This paper introduces a deep residual network and Rosetta protocol for predicting inter-residue orientations and distances, outperforming existing methods in benchmark tests.