Michael P. Dunne

LLPSight predicts LLPS-driving proteins using machine learning and protein language models, achieving an F1 score of 0.885 and identifying promising targets in the human proteome.

This study uses protein language models to rank protein-peptide binding affinities without labeled data or structural information, showing agreement with experimental IC50 values across five datasets.

The paper develops machine learning predictors for amyloid formation using a new database of naturally occurring cross-beta amyloids, showing improved performance over existing methods.

The paper improves a model for predicting protein sequences from structures, achieving higher accuracy and showing sampled sequences fold into similar structures.

This paper introduces a deep learning model, ConvTOX, that classifies protein toxins across all domains of life with high accuracy.

OMGene improves gene models by optimizing evolutionary conservation within orthogroups using RNA-seq data from plants and fungi.

OrthoFiller improves genome annotations by identifying and adding missing genes using information from multiple related species.