| Abstract: |
| Most diseases are characterized by complex dynamical features that emerge from interaction between multi-scale components, including biomolecules, cells, organs, and even patients. Furthermore, current clinical and experimental approaches often only provide limited data that spin between different scales and differ between individuals. Novel theoretical methods are needed to effectively utilize these data to design personalized diagnosis and treatment. In this work, we illustrate how a novel methodology that combines the power of mechanistic modeling, machine learning (ML) and artificial intelligence (AI) could overcome the limitation of each individual tool and extract promising biomarkers that can guide personalized diagnosis and treatment. We expect that the board application of this general methodology could help improve the diagnosis and facilitate the development of novel treatment of many complex diseases. |
|