Many systems in nature, society and technology are composed of numerous nonlinearly interacting parts. The dynamic organization of these systems often allows the emergence of intermediate structures that once formed deeply affect the system, and therefore play a key role in understanding its behavior. An interesting hypothesis is that the simultaneous analysis of a system at both levels of description (the microlevel of the relationships between single entities, and the mesolevel constituted by their dynamically organized groups) can allow a better understanding of the phenomenon under examination. In this work we apply this idea to a cancer evolution model, of which each individual patient represents a particular instance. Specifically, in order to validate the idea we analyze the same synthetic dataset – whose ground truth is known – with two methods of analysis, and we merge the results in an innovative way. In doing this, we also evaluate the effectiveness of a new method of reconstructing networks of relationships.
D'Addese, G., Graudenzi, A., La Rocca, L., Villani, M. (2022). Two-Level Detection of Dynamic Organization in Cancer Evolution Models. In Artificial Life and Evolutionary Computation 15th Italian Workshop, WIVACE 2021, Winterthur, Switzerland, September 15–17, 2021, Revised Selected Papers (pp.207-224). Springer Science and Business Media Deutschland GmbH [10.1007/978-3-031-23929-8_20].
Two-Level Detection of Dynamic Organization in Cancer Evolution Models
Graudenzi A.;
2022
Abstract
Many systems in nature, society and technology are composed of numerous nonlinearly interacting parts. The dynamic organization of these systems often allows the emergence of intermediate structures that once formed deeply affect the system, and therefore play a key role in understanding its behavior. An interesting hypothesis is that the simultaneous analysis of a system at both levels of description (the microlevel of the relationships between single entities, and the mesolevel constituted by their dynamically organized groups) can allow a better understanding of the phenomenon under examination. In this work we apply this idea to a cancer evolution model, of which each individual patient represents a particular instance. Specifically, in order to validate the idea we analyze the same synthetic dataset – whose ground truth is known – with two methods of analysis, and we merge the results in an innovative way. In doing this, we also evaluate the effectiveness of a new method of reconstructing networks of relationships.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
