A study from the University of Padua lays the foundation for a new generation of diagnostic and therapeutic tools that combine network science, computational biology, and digital medicine. The future of personalized medicine is taking on a new shape thanks to a transparent approach based on computational models that overcome the limitations of traditional artificial intelligence methodologies. An international team led by researchers from the University of Padua's Padua Center for Network Medicine has proposed a new conceptual framework for the use of digital twins in precision medicine. A digital twin is a virtual model of a physical object that follows the object's lifecycle and uses real-time data sent by sensors on the object itself to simulate its behavior and monitor operations. The results of the research titled "Challenges and opportunities for digital twins in precision medicine from a complex systems perspective" were published in the scientific journal "NPJ Digital Medicine", part of the Nature publishing group, and lay the foundation for a new generation of diagnostic and therapeutic tools that combine network science, computational biology, and digital medicine.
"Our approach is not a simple predictive modeling exercise," explains Manlio De Domenico, first author of the study and professor at the Department of Physics and Astronomy at the University of Padua. "It is based on complex computational models guided by explicit biological hypotheses, which allow us to simulate and analyze therapeutic interventions in a transparent and interpretable way, improving the understanding of the mechanisms underlying biological processes. The challenge now is to make them communicate well with each other."
An interdisciplinary platform for future medicine
The study stands out for its interdisciplinarity, as it integrates concepts and techniques from statistical physics with biology and medicine. The digital twins described in the work are not simple statistical reproductions of clinical data but actual explanatory models that, in principle, are able to replicate in-silico the behavior of cells, organs, or entire organisms using simulations based on multi-scale and multi-level biological mechanisms. This allows for exploring dynamic therapeutic strategies and optimizing clinical decisions in real-time. The research, conducted in collaboration with Ca' Foscari University of Venice, Binghamton University (USA), the London Institute for Mathematical Sciences, and the Catholic University of Portugal in Lisbon, highlights how these models can fill the gaps of "opaque" artificial intelligence techniques, defined as such because their complexity prevents human users from fully understanding and explaining the mechanisms that guide them, arousing a certain distrust in their use. This complexity hinders the spread of artificial intelligence in crucial sectors such as medicine and security.
"Our goal is to make personalized medicine more reliable and understandable, avoiding the opacity of purely data-driven solutions," adds Valeria d'Andrea, a researcher at the Department of Physics and Astronomy of the University of Padua, who contributed to the study.
Practical impacts and future applications
Thanks to the use of hypothesis-driven generative models, this approach promises to improve the effectiveness of personalized therapies, reducing the risks associated with suboptimal diagnoses and treatments. The integration of biological, historical, and environmental "big data" also allows for capturing the complexity of biological interactions and the exposome (the set of environmental stimuli that come into contact with the body), opening up new possibilities in the fight against complex diseases such as cancer, neurodegenerative diseases, and many chronic pathologies. "Digital twins not only simulate realistic clinical scenarios but allow for testing therapeutic interventions safely and efficiently, providing a spectrum of clinically relevant alternatives to support doctors' decisions," concludes De Domenico. This research represents a meeting point between complex systems physics, medicine, and systems biology and outlines new perspectives for the development of more equitable, effective, and sustainable medicine, marking an important step towards realizing the potential of precision medicine. The commitment of the Padua Center for Network Medicine at the University of Padua in this direction underlines the central role of interdisciplinary research in the transformation of modern medicine.
Link: https://www.nature.com/articles/s41746-024-01402-3
Title: Challenges and opportunities for digital twins in precision medicine from a complex systems perspective – "NPJ Digital Medicine" – 2025
Authors: Manlio De Domenico, Luca Allegri, Guido Caldarelli, Valeria d'Andrea, Barbara Di Camillo, Luis M. Rocha, Jordan Rozum, Riccardo Sbarbati, & Francesco Zambelli
