Filippo Castiglione

This paper develops a multiscale computational framework to simulate drug-host-pathogen interactions and quantifies the relative contributions of drug and immune responses to bacterial clearance in tuberculosis treatment.

This study benchmarks protein language models for predicting protein crystallization propensity and identifies novel potentially crystallizable proteins.

This paper uses computational methods to design a multi-serotype Dengue virus vaccine by predicting epitopes and validating safety, offering a novel approach for broad-spectrum immunity.

MATE-Pred predicts TCR-epitope binding affinity using multi-modal attention, outperforming baselines with +8.4% MCC and +5.5% AUC.

This paper develops an in-silico model to simulate and analyze the immune response to SARS-CoV-2, identifying factors like viral load and antibody titers that influence clinical outcomes.

The paper identifies critical nodes in the human protein-protein interaction network that have the greatest impact on network fragmentation and compares their biological relevance to essential genes.