513225 research outputs found
Sort by
Decoding the Complexity of Nucleic Acids: Computational Approaches Across Multiscale Levels
No full textIn response to the growing interest within the scientific community, there has been an increasing commitment to the study of nucleic acids in recent years. Correlations have been identified between their properties and several pathological conditions, and there is considerable promise for the utilization of these systems in innovative pharmaceutical applications. In light of the contemporary scientific significance and the state of the art of computational methods for these biopolymers, the present thesis endeavors to address the challenges in the theoretical modeling of such systems using a multiscale approach. Initially, insights concerning the hydrolysis reaction were obtained from model molecules describing the backbone of nucleic acids. Subsequently, an investigation was conducted into the stability of nucleobases to assess the efficacy of computational approaches in predicting genome damage. These results provide a basis for evaluating the efficacy of quantum mechanics methods. Finally, alchemical free energy calculations were used to characterize a complete sequence binding behavior. Consequently, this study is a notable instance of effective nucleic acid modeling through the application of molecular mechanics. The preceding theoretical framework enabled the investigation of a protein-nucleic acid complex. The system was thoroughly characterized, and the resulting data was capable of elucidating subtle mechanisms implicated in the onset and progression of diseases and aging. Additionally, the role of machine learning in this context was investigated by assessing the capability of a recently developed machine learning code in predicting the folding of a short sequence. This thesis constitutes a comprehensive study of the current level of achievement in the field of computational modeling of nucleic acids. Through these efforts, the effectiveness of these approaches in the characterization of the processes involving these molecules was substantiated
Profili giuridici e tributari dei non-fungible token (nft): analisi e prospettive di riforma
No full textLa tesi analizza in modo sistematico la fiscalità dei Non-Fungible Token (NFT), affrontandone la natura giuridica e le implicazioni tributarie dirette e indirette, nel contesto del diritto italiano, europeo e internazionale. Gli NFT, certificati digitali unici basati su tecnologia blockchain, pongono questioni di qualificazione che incidono sui principi costituzionali di legalità e capacità contributiva. L’assenza di una disciplina normativa espressa ha generato incertezze applicative e un ricorso eccessivo a interpretazioni analogiche, con il rischio di violare la riserva di legge in materia fiscale. La ricerca propone un inquadramento organico degli NFT, fondato sul principio di prevalenza della sostanza sulla forma, valorizzando la funzione economico-giuridica del token. Dopo aver esaminato la disciplina dell’imposizione diretta e indiretta, la fiscalità dei marketplace e i profili comparati (OCSE, MiCA, DAC8, CARF), la tesi giunge a sostenere la necessità di un intervento legislativo che distingua gli NFT dalle criptovalute, introducendo una normativa specifica capace di garantire certezza del diritto, neutralità fiscale e coerenza sistematica nell’era digitale
Mechanics of swelling
No full textSoft materials are all around us: they naturally form biological tissues and plant matter, while rubbers and gels are used as sealants, contact lenses, energy absorbers, sensors, and actuators. Most soft materials swell as they absorb fluids – in some cases this swelling is a desirable feature of the material, while in others it can be detrimental to its ability to function. It is essential to understand and model how soft materials swell, and how these swollen rubbers and gels respond to the forces they experience.
This book combines perspectives from soft matter physics, nonlinear mechanics, and materials science to describe the mechanics of swelling, and serves as a reference for anyone interested in the mechanics of swelling in soft polymers, regardless of their scientific background. It provides the necessary physical and mechanical knowledge to model the behaviour of soft materials that swell by offering a comprehensive exploration of theoretical concepts, complemented by practical computational exercises
Strategia di ottimizzazione del monitoraggio e controllo ambientale degli edifici basata sul digital twin
No full textNell’attuale scenario del settore dell'architettura, dell'ingegneria e delle costruzioni (AEC), la transizione verso paradigmi di gestione edilizia intelligente e sostenibile si configura come un imperativo strategico. In questo contesto, l'introduzione dei Digital Twin (DT) rappresenta una metodologia all'avanguardia per l'ottimizzazione delle prestazioni operative degli edifici, offrendo strumenti innovativi per il monitoraggio e il controllo. Il presente studio si inserisce in questa cornice, esplorando l'applicazione di tali principi in un caso di studio complesso, con l'ambizione di stabilire un modello replicabile che possa estendersi a futuri sviluppi, come quello del progetto Rome Technopole.
Il presente studio analizza lo sviluppo e la validazione di un prototipo di Digital Twin, denominato QTwin, applicato alla gestione e al monitoraggio avanzato di alcuni ambienti di un edificio storico e istituzionale situato a Roma. Questo studio mira a perseguire molteplici obiettivi, tra cui migliorare l'efficienza energetica e la qualità dell'aria interna, promuovendo al contempo la manutenzione predittiva e ottimizzando l'utilizzo degli spazi. La ricerca affronta intrinseche complessità legate all'integrazione tecnologica in un contesto edilizio vincolato, caratterizzato da ostacoli alla connettività wireless e dalla necessità di aderire a rigorosi standard di cybersecurity e interoperabilità, aspetto per il quale la letteratura accademica risulta ancora carente.
La metodologia adottata ha previsto una selezione e valutazione sistemica delle piattaforme per lo sviluppo di Digital Twin, dei dispositivi Internet of Things (IoT) e dei loro protocolli di comunicazione. È stata implementata un'architettura di sistema basata sul framework sviluppato nell’ambito del progetto di ricerca SmartLab, un ambiente dimostrativo sperimentale situato all'interno della Facoltà di Architettura della Sapienza Università di Roma. Questa architettura integra una vasta rete di sensori IoT per la raccolta in tempo reale di dati essenziali, quali consumi energetici, qualità dell'aria interna (IAQ), comfort termico e livelli di occupazione degli spazi.
Questi dati sono stati trasmessi e gestiti attraverso protocolli standard e processati da un'infrastruttura di backend che include sistemi di archiviazione e visualizzazione tramite dashboard interattive. Per l'interpretazione dei dati, la formulazione di previsioni e l'identificazione di anomalie, sono stati utilizzati algoritmi avanzati di Machine Learning (ML), con un'elaborazione che avviene anche a livello locale (edge computing) per assicurare risposte tempestive. Il sistema include inoltre un meccanismo "Human-in-the-loop" per le azioni che richiedono l'intervento umano, ed è stato rigorosamente validato attraverso il confronto tra dati reali e simulazioni predittive, oltre a stress test.
I risultati ottenuti evidenziano l'efficacia dell'integrazione tra le tecnologie IoT, i sistemi di automazione e gli algoritmi di apprendimento automatico all'interno della piattaforma. Questa sinergia consente una rappresentazione dinamica e in tempo reale dei dati, abilitata dalla visualizzazione di modelli grafici e informativi sviluppati secondo la metodologia Building Information Modeling (BIM) e dall'utilizzo di cruscotti interattivi.
I modelli di Machine Learning sviluppati per la previsione del consumo energetico hanno mostrato un'elevata accuratezza, riuscendo a identificare sia le tendenze generali che le fluttuazioni stagionali. L'analisi della concentrazione di anidride carbonica (CO2) ha permesso di stimare l'occupazione e di identificare periodi di qualità dell'aria interna non ottimale, in linea con le normative vigenti. Il monitoraggio della temperatura (°C) e dell'umidità relativa (UR) ha fornito dati essenziali per prendere decisioni informate sul comfort termico e sull'efficienza, mentre tecniche di analisi dei dati hanno permesso di identificare i pattern di consumo, promuovendo strategie di gestione energetica più consapevoli.
Questi esiti attestano la validità e la replicabilità del framework in contesti edilizi complessi ed eterogenei, evidenziando la scalabilità e l'adattabilità dell'approccio Digital Twin per una gestione edilizia sostenibile e intelligente. Il progetto QTwin offre una comprensione approfondita e strumenti decisionali cruciali per una gestione tempestiva data-driven, prefigurando la possibilità di estendere questa metodologia ad altri ambiti urbani o infrastrutturali. Le future direzioni di ricerca includono il perfezionamento dei modelli di apprendimento automatico, l'ottimizzazione dell'interazione uomo-macchina, un approfondimento sulla cybersecurity e la standardizzazione per una maggiore interoperabilità
Lightweight Anomaly Detection for IoT: Evaluating Machine Learning and Deep Learning Models on CICIDS2017
No full textGiven the increasing rate of cyber attacks, specifically Denial of Service (DoS) attacks, there is a growing need for fast and efficient Intrusion Detection Systems (IDS). In this work, we studied the implementation of real-time IDS within resource constrained environments like Internet of Things (IoT) networks. We studied and tested a wide range of Machine Learning and Deep Learning models applied to the CICIDS2017 dataset, a commonly used benchmarking tool for network intrusion detection. We compared the results of models such as Logistic Regression, Random Forest, XGBoost, K-Nearest Neighbors, Support Vector Machines, Single-layer Perceptron (SLP), Multi-layer Perceptron (MLP), Deep Convolutional Neural Network (DCNN), ResNet, and DenseNet. We focused our investigation on performance metrics such as accuracy, precision, recall, F1-score, and inference time, trying to find the model with the best trade-off between detection capability and computation overhead considering the constrained resources of IoT devices. The results highlight that real-time security of IoT infrastructures with minimal resource consumption is possible with simple models such as XGBoost, SLP, or MLP
L’aquila di Esopo parla al drago cinese: il caso dello Yishi yuyan 意拾喻言
No full textQuesto studio esplora il complesso processo di ricezione e diffusione delle Favole di Esopo in Cina attraverso la lente della comunicazione interculturale. Attraverso un’analisi filologica e storico-culturale delle fonti, in particolare documenti redatti da missionari e mercanti occidentali durante le dinastie Ming e Qing, la ricerca esamina le caratteristiche distintive delle diverse fasi della sua trasmissione. Il focus si concentra sulla traduzione Yishi yuyan 意拾喻言 (1840), realizzata dal britannico Robert Thom in collaborazione con il suo assistente cinese, che viene analizzata come caso studio esemplare delle strategie di adattamento culturale. Questa opera, vero punto di svolta nella storia delle traduzioni cinesi di Esopo, mostra come un classico possa essere rielaborato in modo creativo e intelligente, bilanciando fedeltà al testo originale e sensibilità al contesto culturale d’arrivo
Helpers’ well-being and workload: from expectations to reality. The development of a new Tool
No full textThe global mental health treatment gap constitutes one of the most significant public health challenges
of the twenty-first century, particularly in low- and middle-income and humanitarian settings. Task-
sharing approaches, where trained non-specialists deliver evidence-based psychosocial interventions,
represent a key strategy to bridge this gap. However, the psychological well-being of the helpers, who
often work under conditions of chronic stress, insecurity, and limited supervision, has received minimal
empirical attention despite their crucial role in intervention sustainability. This doctoral research
addresses this gap by examining the determinants, experiences, and measurement of helpers’ well-being
across diverse humanitarian and low-resource contexts, and by developing a novel assessment tool for
this purpose.
The project integrates theoretical perspectives from ecological systems theory, occupational health
psychology, and implementation science, and comprises six interrelated studies employing systematic,
qualitative, mixed-method, and psychometric methodologies. A systematic review of a WHO’s scalable
intervention established the evidence base for implementation outcomes, revealing strong feasibility and
acceptability but limited data on sustainability and workforce support. Subsequent qualitative and
mixed-method studies conducted in Haiti, Sub-Saharan Africa, Eastern Europe, and Europe explored
contextual barriers and facilitators, confirming that helpers’ well-being is both a determinant and a
consequence of effective intervention delivery. Across studies, supportive supervision, balanced
workload, and organisational recognition emerged as critical to sustaining motivation and fidelity, while
a recurrent provider support paradox was identified: those responsible for supporting others often
receive insufficient emotional and organisational support themselves. Building upon these findings, the
Well-being and Workload Assessment Tool (WWAT) was developed and psychometrically validated to
provide a reliable, contextually grounded measure of helper well-being, workload, and support systems.
The WWAT demonstrated strong internal consistency, a four-factor structure, and theoretical coherence
with established constructs, offering a practical means for ongoing workforce monitoring.
This thesis contributes to global mental health and implementation science by conceptualising helper
well-being as a core implementation outcome essential to ethical, effective, and sustainable delivery of
scalable psychological interventions. It advances a multi-level, systems-oriented model linking
individual, organisational, and systemic determinants of well-being to programme quality and
sustainability. The identification of the provider support paradox further underscores the need to embed
structured mechanisms of supervision, recognition, and psychological support within humanitarian and
public health systems, ensuring that those who provide care are themselves adequately cared for
GGS: Generations and Gender Survey
No full textThe Generations and Gender Programme (GGP) is a research infrastructure that generates harmonised, high-quality, and timely data on families and individual life-course trajectories. Its flagship instrument—the Generations and Gender Survey (GGS)—provides comparable evidence on demographic behaviour, intergenerational relationships, and the evolving social roles of women and men, thereby enabling scholars and policy-makers to address pressing societal challenges. Currently, a second round of comparative data collection is ongoing and will deliver fresh insights into persistent low fertility, the increasingly complex transition to adulthood, and new forms of family solidarity. In Italy, Wave 1 of the second round started in summer 2025, supported by the PNRR projects “Fostering Open Science in Social Science Research” (FOSSR) and “Aging Well in an Aging Society” (Age-It)
High Habitat Potential but Limited Connectivity for Brown Bears Throughout Europe
No full textAim: Large carnivores worldwide have experienced substantial range contractions due to human activities, though several species are recolonising parts of their historical range. We aimed to assess current and potential European brown bear (Ursus arctos arctos) habitat as well as habitat connectivity on a continental scale.
Location: The extended biogeographical regions of Europe, spanning from Portugal to central Russia, longitudinally, and from Norway to Türkiye, latitudinally. Excluding inland seas; this area covers 11,151,636 km2.
Methods: We assessed habitat suitability throughout the study area using an ensemble species distribution model with nine submodels, using data from 10 European bear populations and Türkiye. We used the resulting habitat suitability maps to conduct a least-cost path connectivity analysis and an omnidirectional circuit connectivity analysis.
Main Conclusions: Habitat suitability was strongly associated with low percentages of agricultural cover, low percentages of human development, and proximity to forest. Of our entire study area, 37% (4.09 million km2) is occupied or potentially suitable for bears.
Connectivity analyses identified corridors that could facilitate movement among southern European bear populations, though agricultural and and human development limit connectivity between northern and southern European bear populations. Previous research estimated bears occupied 0.5 million km2 across the European Union, while our results estimate 1.82 million km2 of this part of our study area is potentially suitable for bears, though connectivity is limited. Our results inform conservation strategies and policy development for the future of brown bears in Europe, emphasising the need for transboundary conservation efforts
Comprehensive genomic profiling on tissue and liquid biopsy for actionable mutations in advanced solid tumors
No full textBackground: Precision oncology has transformed the treatment landscape for advanced solid tumors through comprehensive genomic profiling (CGP). While the ROME trial demonstrated clinical superiority of CGP-guided targeted therapy over standard of care (SoC), the impact of concordance between tissue and liquid biopsy on clinical outcomes remains poorly defined. This study aims to evaluate how concordance between these two biopsy modalities influences the efficacy of targeted therapy (TT) in patients with pretreated advanced solid tumors.
Methods: Genomic data from 400 patients enrolled in the ROME trial who received both tissue
biopsy (FoundationOne CDx) and liquid biopsy (FoundationOne Liquid CDx) were retrospectively analyzed. Patients were stratified into 3 groups: concordant (T+L, alterations detected in both modalities, n=197), tissue-only (T, n=139), and liquid-only (L, n=64). Primary endpoints included objective response rate (ORR), progression-free survival (PFS), and overall survival (OS). Subgroup analyses evaluated the impact of tumor fraction, number of metastatic sites, and primary tumor type on concordance.
Results: Concordance between tissue and liquid biopsy was observed in 49.2% of patients. In the concordant group, targeted therapy demonstrated significant superiority over SoC with an ORR of 22.1% versus 11.8% (p=0.042), median PFS of 4.90 versus 2.80 months (HR 0.56, 95% CI 0.41-0.77), and median OS of 10.89 versus 7.42 months (HR 0.67, 95% CI 0.48-0.95). Among patients receiving targeted therapy, outcomes revealed that the T+L group achieved the longest median PFS (4.90 months), followed by the T group (3.06 months) and the L group (2.07 months). Similar patterns were observed for OS, with medians of 10.89, 9.93, and 4.05 months, respectively. Discordance was primarily attributed to differences in molecular alterations (43%), TMB divergences (35%), and technical failures (21%). High tumor fraction (>10%) emerged as a significant predictor of concordance (p<0.001). The PTEN/PI3K/AKT/mTOR pathway exhibited the highest discordance rate (50.5%), followed by FGF/FGFR (15.4%) and ERBB2 (13.2%).
Conclusions: Concordance between tissue and liquid biopsy identifies patients with higher likelihood of benefiting from CGP-guided targeted therapy. These findings support the integration of both biopsy modalities in clinical practice when feasible. Discordance may reflect tumor heterogeneity, clonal evolution, or technical limitations, underscoring the need for personalized and integrated biopsy strategies. Future studies should explore strategies to optimize the combined use of both modalities and identify patient subgroups that may derive the greatest benefit from dual-biopsy approaches