Ludovica Beltrame

This paper predicts antibody non-specificity using protein language models and biophysical parameters, showing that heavy variable domains and isoelectric point are key predictors.

MAVISp provides a modular structural framework for analyzing protein variant effects, offering over eight million variants across 1000 proteins with curated data and a web server for accessibility.

The paper uses MAVISp and molecular dynamics simulations to analyze DNA polymerase variants, classifying VUS and predicting functional consequences of POLE and POLD1 variants.

PDBminer provides an open-source tool to search and rank experimentally determined protein structures from PDB and AlphaFoldDB for computational analysis.

The paper introduces two open-source tools for analyzing protein structural ensembles using network theory, focusing on reproducibility and parameter stability.

RosettaDDGPrediction enables high-throughput calculation of free energy changes from amino acid substitutions using Rosetta, with tools for data aggregation and visualization.

MutateX automates in-silico saturation mutagenesis to predict free energy changes of protein mutations, supporting analysis of disease-associated mutations and protein design.