Alessandro Iannitti
Advanced Molecular Dynamics Simulations of Cyclic Peptide-Protein Complexes.
Rel. Carla Bosia, Dirk Stratmann, Fabio Pietrucci. Politecnico di Torino, Corso di laurea magistrale in Physics Of Complex Systems (Fisica Dei Sistemi Complessi), 2026
|
Preview |
PDF (Tesi_di_laurea)
- Tesi
Licenza: Creative Commons Attribution Non-commercial No Derivatives. Download (18MB) | Preview |
Abstract
Cyclic peptides are an important class of pharmaceutical drugs, offering an optimal balance of high affinity, specificity and metabolic stability. Their computational modeling remains, however, challenging due to the complex interplay between conformational flexibility and structural rigidity. In this thesis we investigate the binding-unbinding mechanisms of cyclic peptide-protein complexes to accurately predict their absolute binding free energies (ΔG) using advanced molecular dynamics (MD) simulations. We focused primarily on the Kelch domain of human KEAP1 bound to different Nrf2-based cyclic peptides (these complexes are called 7K2*) and then validated our approach on a highly rigid and larger system (1SFI). In order to overcome the sampling limitations of classical MD, we combined a Hamiltonian Replica Exchange (HREX) setup driven by On-the-fly Probability Enhanced Sampling (OneOPES) with Funnel Metadynamics, efficiently biasing hydration collective variables, while restraining peptide exploration in the unbound state.
We have tested two force fields, Amber14SB and a99SB-disp, for each complex, finding a profound structural bias: classical force fields like Amber14SB work really well on rigid systems but over-compact flexible peptides in the unbound state, thus overestimating the binding free energy
Relatori
Anno Accademico
Tipo di pubblicazione
Numero di pagine
Corso di laurea
Classe di laurea
Aziende collaboratrici
URI
![]() |
Modifica (riservato agli operatori) |
