Frances H. Arnold

This paper introduces AlphaProtein Novo, a machine-learning pipeline for de novo enzyme design that outperforms natural sequence mining in creating enzymes for new-to-nature chemistry and substrates.

This study combines design and evolution to create new-to-nature enzymes capable of performing regio- and enantioselective C-H amination reactions for synthesizing chiral piperidines.

EZSolver predicts polar enzymatic mechanisms without templates using flow matching and beam search, achieving 60% accuracy on unseen reactions.

This paper uses machine learning to guide the creation of enzyme libraries with broader catalytic activity, showing improved performance across multiple reactions.

This paper introduces Enzyme-toolkit, a framework for discovering new enzymes with specific activities, demonstrating its effectiveness in identifying enzymes for pollutant degradation.

This paper uses a generative AI model to design novel tryptophan synthase variants with enhanced catalytic activity and substrate promiscuity, outperforming natural and optimized versions.

Squidly predicts catalytic residues using a biology-informed contrastive learning framework, outperforming existing methods and integrating with BLAST for efficient annotation.

This paper introduces ALDE, a machine learning-assisted directed evolution method that improves protein engineering outcomes by efficiently exploring epistatic mutation spaces.

LevSeq rapidly generates sequence-function data for protein engineering using nanopore sequencing and dual barcoding, enabling directed evolution and machine learning applications.

This study evaluates machine learning-assisted directed evolution strategies across diverse protein fitness landscapes, finding that focused training with zero-shot predictors outperforms random sampling in challenging landscapes.

The paper constructs a large-scale sequence-fitness dataset for enzyme engineering, revealing significant epistasis and highlighting challenges in predicting optimal variants.

This paper introduces MLDE, a machine learning-assisted directed evolution method that outperforms traditional greedy optimization in navigating combinatorial fitness landscapes with reduced screening burden.

This paper develops a continuous directed evolution platform called OrthoRep to generate diverse enzyme variants with promiscuous activities by evolving TmTrpB through multi-mutation pathways.