From Bioinformatics to Systems Biology

Mining bioinformatic data is key to identify the salient genes involved in pathophysiologic situations at hand. In particular Logical Clustering (Muselli and Liberati IEEE Transactions on Knowledge and data Engineering 2002) also allows to infer understandable rules linking the data. In simplified cases, like discriminating myeloid from linmhatic leukemia (Garatti et al., Intelligent Data Analysis 2007), a principal divisive approach based on principal components is sufficient, as today exemplified also in Clustering of Pancreatic Endocrine Tumors via Microarray Gene Expression Analysis, the first of the two papers introduced in this lecture. Analyzing dynamics, a more complete approach does allow for instance to model the biochemistry of the interplay among Ras and SoS (Sacco, Farina et al. Biotechnology Advance, 2012) as well as the biophysical-biochemical effect of the tumor necrosis factor in modulating apoptosis especially in bystander effect in radiotherapy, as also illustrated in the second paper presented today: A Quantitative Numerical Model for TNF-? Mediated Cellular Apoptosis.