Matteo De Leonardis

The paper models protein sequence evolution using an Ornstein-Uhlenbeck process in a latent space to improve contact prediction by integrating directed evolution data with natural homologs.

This paper introduces a method using autoregressive generative models to reconstruct ancestral protein sequences while accounting for epistasis, outperforming existing methods in simulations and experimental data.

This paper introduces a biophysically-inspired model to predict the viability of AAV2 capsid variants using DMS data, evaluating its performance on log-selectivity and binary classification.