Quantum computing is an emerging technology that offers, for specific classes of problems, a potential computational advantage over classical methods. The main current challenges concern overcoming hardware limitations and developing more effective algorithms. Certain computational paradigms, in particular quantum annealing and hybrid approaches that integrate quantum and classical resources, are progressively moving beyond the purely academic context. Since the benefits are confined to well-defined problem types, it is essential to accurately identify the characteristics of problems that can benefit from quantum computing and to map them onto real-world applications. In this study, we focus on the quantum annealing paradigm and identify the protein side-chain optimization problem as the case study. Given the backbone structure of a protein, this problem consists of determining the optimal configuration of rotamers for each residue in order to minimize the overall energy of the macromolecule. Initially a theoretical study of the characteristics of problems suited to annealing and their alignment with the selected case study is developed, considering also the contextualization of the problem within an applied setting in the field of protein engineering. To assess the potential computational advantage, a model is developed in three variants and implemented using Quantum Annealing, Hybrid Annealing, and the commercial solver Gurobi, which serves as a reference for validation and benchmarking. The experiments, conducted on real protein structures and realistic problem sizes, identify the hybrid solver as the most promising solution, yielding encouraging results while highlighting the need for further analysis across different benchmarks and hardware platforms.
Il calcolo quantistico è una tecnologia emergente che offre, per specifiche classi di problemi, un potenziale vantaggio computazionale rispetto ai metodi classici. Le principali sfide attuali riguardano il superamento dei limiti hardware e lo sviluppo di algoritmi più efficaci. Alcune modalità di calcolo, in particolare il quantum annealing e gli approcci ibridi che integrano risorse quantistiche e classiche, stanno progressivamente superando il contesto puramente accademico. Poiché i benefici sono circoscritti a tipologie di problemi ben definite, risulta fondamentale identificare con precisione le caratteristiche dei problemi che possono trarre vantaggio dal calcolo quantistico e mappare tali problemi su applicazioni reali. In questo studio ci concentriamo sul paradigma del quantum annealing e individuiamo nel problema dell’ottimizzazione delle catene laterali delle proteine il caso di studio. Tale problema consiste, data la struttura del backbone di una proteina, nell’identificare la configurazione ottimale dei rotameri per ciascun residuo, al fine di minimizzare l’energia complessiva della macromolecola. Inizialmente, viene sviluppato uno studio teorico delle caratteristiche dei problemi adatti al quantum annealing e del loro allineamento con il caso di studio selezionato, considerando anche la contestualizzazione del problema in un contesto applicativo nel campo dell’ingegneria proteica. Per verificare l’eventuale vantaggio computazionale, viene quindi sviluppato un modello in tre varianti, implementato mediante Quantum Annealing, Annealing Ibrido e il solver commerciale Gurobi, utilizzato come riferimento per la verifica e il benchmarking. I test, condotti su strutture proteiche reali e su dimensioni realistiche, individuano nel solver ibrido la soluzione più promettente, mostrando risultati incoraggianti, pur evidenziando la necessità di ulteriori analisi su benchmark e piattaforme hardware differenti.
Perspectives on quantum computing applications via hybrid quantum annealing in protein engineering
Visani, Valeria Benedetta Cecilia
2024/2025
Abstract
Quantum computing is an emerging technology that offers, for specific classes of problems, a potential computational advantage over classical methods. The main current challenges concern overcoming hardware limitations and developing more effective algorithms. Certain computational paradigms, in particular quantum annealing and hybrid approaches that integrate quantum and classical resources, are progressively moving beyond the purely academic context. Since the benefits are confined to well-defined problem types, it is essential to accurately identify the characteristics of problems that can benefit from quantum computing and to map them onto real-world applications. In this study, we focus on the quantum annealing paradigm and identify the protein side-chain optimization problem as the case study. Given the backbone structure of a protein, this problem consists of determining the optimal configuration of rotamers for each residue in order to minimize the overall energy of the macromolecule. Initially a theoretical study of the characteristics of problems suited to annealing and their alignment with the selected case study is developed, considering also the contextualization of the problem within an applied setting in the field of protein engineering. To assess the potential computational advantage, a model is developed in three variants and implemented using Quantum Annealing, Hybrid Annealing, and the commercial solver Gurobi, which serves as a reference for validation and benchmarking. The experiments, conducted on real protein structures and realistic problem sizes, identify the hybrid solver as the most promising solution, yielding encouraging results while highlighting the need for further analysis across different benchmarks and hardware platforms.| File | Dimensione | Formato | |
|---|---|---|---|
|
2026_03_Visani_Executive Summary.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Executive Summary
Dimensione
612.03 kB
Formato
Adobe PDF
|
612.03 kB | Adobe PDF | Visualizza/Apri |
|
2026_03_Visani_Tesi.pdf
non accessibile
Descrizione: Tesi
Dimensione
6.69 MB
Formato
Adobe PDF
|
6.69 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/252496