Jerry M. Parks

Benchmarking AI Protein Structure Predictors Reveals a Persistent Bias in Multi-State Proteins

This study benchmarks AI protein structure predictors on multi-state proteins, revealing persistent biases in capturing conformational ensembles and ligand interactions.

Structural and oligomeric characterization of substrate- and product-selective nylon hydrolases

This study characterizes three nylon hydrolases, identifying structural features and activity differences that inform their substrate specificity and potential for engineering.

IRIS: A Machine Learning-Based Pose Re-Ranking Tool for RNA-Ligand Docking

IRIS improves RNA-ligand docking pose ranking by integrating a regression model trained on experimental complexes, increasing accuracy in selecting near-native poses.

Improved nylon polymerization using amide diads

This paper demonstrates a method to produce higher-molecular-weight nylons using chemically-synthesized diads and biosynthetic amide synthetases, expanding the range of usable precursors.

Identification and characterization of substrate- and product-selective nylon hydrolases

The paper identifies and characterizes diverse nylon hydrolases, showing that 40% of tested enzymes have significant activity, with some showing substrate selectivity for Nylon 6,6 over Nylon 6.

Predicting direct physical interactions in multimeric proteins with deep learning

The paper adapts AlphaFold2 models to predict structures of multimeric protein complexes and introduces metrics for predicting direct protein-protein interactions.

Hotspot coevolution at protein-protein interfaces is a key identifier of native protein complexes

This study introduces KFC-E, a scoring function that uses conservation and coevolution of hotspot residues to identify native protein complexes, achieving better performance than alternative methods.