A. M. Watkins

Rapid generation of antibody conformational ensembles using machine learning

This paper introduces Cadense, a machine learning framework that generates antibody conformational ensembles from sequence, capturing dynamic properties and improving docking accuracy.

Local structural flexibility drives oligomorphism in computationally designed protein assemblies

This study shows that structural flexibility in computationally designed protein assemblies leads to multiple architectures, expanding their functional possibilities.

Mixed structure- and sequence-based approach for protein graph neural networks with application to antibody developability prediction

This paper introduces a graph neural network for antibody developability prediction using structure- and sequence-based methods, achieving state-of-the-art performance.

RNA 3D modeling with FARFAR2, online

The paper demonstrates the use of FARFAR2 for RNA 3D modeling with two case studies and experimental data integration.

Theoretical basis for stabilizing messenger RNA through secondary structure design

This paper develops a biophysical model to predict RNA stability and demonstrates that superfolder mRNAs can increase in vitro half-life by at least two-fold through AUP optimization.

De novo 3D models of SARS-CoV-2 RNA elements and small- molecule-binding RNAs to aid drug discovery

The paper provides 3D models of SARS-CoV-2 RNA elements and small-molecule-binding RNAs to support drug discovery efforts using computational modeling and experimental validation.

Designing Peptides on a Quantum Computer

This paper maps protein design to a quantum annealer, demonstrating real-world applicability of quantum computing in designing peptides comparable to classical methods.

FARFAR2: Improved de novo Rosetta prediction of complex global RNA folds

The paper improves RNA 3D structure prediction using FARFAR2, achieving better accuracy than previous methods in RNA-Puzzles trials without experimental data or expert intervention.

Ribosolve: Rapid determination of three-dimensional RNA-only structures

Ribosolve rapidly determines 3D RNA-only structures using cryo-EM, chemical mapping, and Rosetta modeling, validated by multiple benchmarks and showing high accuracy.

Blind prediction of noncanonical RNA structure at atomic accuracy

This paper introduces a new method for predicting noncanonical RNA structures with high accuracy, validated through benchmarking and blind challenges.