Proteins sit at the core of biological adaptation and disease, yet connecting genetic sequence variation to pathophysiological molecular mechanisms remains a major challenge. This thesis develops and applies integrative computational strategies to identify how proteins encode function, how that function is rewired by evolution or pathology, and how such knowledge can be leveraged to design selective therapeutic or mechanistic tools. Across diverse biological contexts, I combine molecular modeling, classical molecular dynamics, enhanced-sampling simulations, protein-protein and protein-DNA docking, evolutionary genomics, as well as quantitative bioinformatics to derive hypotheses that are experimentally testable and, where possible, translationally relevant. The computational framework established here emphasizes three pillars. First, structural and dynamical modeling is used to resolve transient conformational states, binding interfaces, and allosteric couplings that are inaccessible to static structures alone. Second, comparative and population-scale sequence analyses are integrated with structure to identify conserved domains, lineage-specific innovations, and pathogenic deviations, enabling interpretations of genotype-phenotype relationships. Third, structure-guided design, spanning peptide, miniprotein, and small-molecule modalities, is employed to perturb selected interfaces with isoform or target specificity, providing functional dissection and, in some cases, prototypes for therapeutic development. Collectively, the work shows how multi-scale computation can bridge the gap from sequence to mechanism to intervention. By unifying physical simulation with evolutionary logic and rational design, this thesis offers a generalized roadmap for studying protein-driven processes in both adaptation and disease, and highlights the role of computational approaches not as an accessory, but as a primary engine of biological discovery and innovation.
Le proteine giocano un ruolo chiave sia nell’adattamento che nelle patologie, perciò collegare variazioni nella sequenza genetica ai meccanismi molecolari patofisiologici rimane una sfida fondamentale. Questa tesi sviluppa e applica strategie computazionali integrative per identificare come le proteine codifichino la funzione, come tale funzione venga rimodulata dall’evoluzione o dalla patologia e come queste conoscenze possano essere sfruttate per progettare strumenti terapeutici o meccanicistici selettivi. In contesti biologici eterogenei, combino modellistica molecolare, dinamica molecolare classica, simulazioni a campionamento avanzato, docking proteina-proteina e proteina-DNA, genomica evolutiva e bioinformatica quantitativa per derivare ipotesi sperimentalmente testabili e, ove possibile, rilevanti dal punto di vista traslazionale. Il framework computazionale qui sviluppato si fonda su tre pilastri. Primo, la modellistica strutturale e dinamica viene utilizzata per risolvere stati conformazionali transitori, interfacce di legame e accoppiamenti allosterici non accessibili alle sole strutture statiche. Secondo, analisi di sequenza comparative e su scala di popolazione sono integrate con l’informazione strutturale per identificare domini conservati, innovazioni specifiche di linea evolutiva e deviazioni patogeniche, consentendo interpretazioni delle relazioni genotipo–fenotipo. Terzo, la progettazione guidata dalla struttura che comprende peptidi, miniproteine e piccole molecole, è impiegata per perturbare interfacce selezionate con specificità di isoforma o di bersaglio, fornendo dissezioni funzionali e, in alcuni casi, prototipi per lo sviluppo terapeutico. Nel complesso, questo lavoro mostra come il calcolo multi-scala possa colmare il divario tra sequenza, meccanismo e intervento. Unificando simulazione fisica, logica evolutiva e design razionale, questa tesi propone una roadmap generalizzabile per lo studio dei processi proteina-mediati nell’adattamento e nella malattia, e mette in evidenza il ruolo degli approcci computazionali non come accessorio, ma come motore primario della scoperta e dell’innovazione in biologia.
Integrative computational approaches to study protein-driven mechanisms in adaptation and disease / Perta, Nunzio. - (2026 Mar 06).
Integrative computational approaches to study protein-driven mechanisms in adaptation and disease
PERTA, NUNZIO
2026-03-06
Abstract
Proteins sit at the core of biological adaptation and disease, yet connecting genetic sequence variation to pathophysiological molecular mechanisms remains a major challenge. This thesis develops and applies integrative computational strategies to identify how proteins encode function, how that function is rewired by evolution or pathology, and how such knowledge can be leveraged to design selective therapeutic or mechanistic tools. Across diverse biological contexts, I combine molecular modeling, classical molecular dynamics, enhanced-sampling simulations, protein-protein and protein-DNA docking, evolutionary genomics, as well as quantitative bioinformatics to derive hypotheses that are experimentally testable and, where possible, translationally relevant. The computational framework established here emphasizes three pillars. First, structural and dynamical modeling is used to resolve transient conformational states, binding interfaces, and allosteric couplings that are inaccessible to static structures alone. Second, comparative and population-scale sequence analyses are integrated with structure to identify conserved domains, lineage-specific innovations, and pathogenic deviations, enabling interpretations of genotype-phenotype relationships. Third, structure-guided design, spanning peptide, miniprotein, and small-molecule modalities, is employed to perturb selected interfaces with isoform or target specificity, providing functional dissection and, in some cases, prototypes for therapeutic development. Collectively, the work shows how multi-scale computation can bridge the gap from sequence to mechanism to intervention. By unifying physical simulation with evolutionary logic and rational design, this thesis offers a generalized roadmap for studying protein-driven processes in both adaptation and disease, and highlights the role of computational approaches not as an accessory, but as a primary engine of biological discovery and innovation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


