Data integration, systems approach and multilevel description of complex biosystems

Abstract

Abstract. Recent years have witnessed the development of new quantitative approaches and theoretical tenets in the biological sciences. The advent of high throughput experiments in genomics, proteomics and electrophysiology (to cite just a few examples) have provided the researchers with unprecedented amounts of data to be analyzed. Large datasets, however can not provide the means to achieve a complete understanding of the underlying biological phenomena, unless they are supplied with a solid theoretical framework and with proper analytical tools. It is now widely accepted that by using and extending some of the paradigmatic principles of what has been called complex systems theory, some degree of advance in this direction can be attained. We will be presenting ways in which by using data integration techniques (linear, non-linear, combinatorial, graphical), multidimensional-multilevel descriptions (multifractal modeling, dimensionality reduction, computational learning), as well as an approach based in systems theory (interaction maps, probabilistic graphical models, non-equilibrium physics) have allowed us to better understand some problems in the interface of Statistical Physics and Computational Biology.

Evolutionary Developmental Biology (Evo‐Devo) Research in Latin America

“…Gene network models, which provide a valuable framework to postulate precise mechanisms and events underlying phenotypic change, have been developed mostly from a qualitative view, but have also included large data sets from model genetic organisms and experimental evidence to build and validate the models and postulate their role in evolution of form (Espinosa‐Soto, ; Alvarez‐Buylla et al., ; Hernández‐Lemus, ). Indeed, one of the conclusions of these modeling efforts is that even qualitative gene regulatory network models are able to reproduce the overall regulatory logic of some differentiation and developmental events, suggesting that the global dynamic of these processes largely depends on the network architecture and nature of the interactions, rather than on particular kinetic details.…”

Mechanistic-enriched models: integrating transcription factor networks and metabolic deregulation in cancer

“…Enrichment implies two things: the first one is a hierarchical perspective. As it has been already pointed out by others, Systems Biology and complex biological systems imply a multilevel organization [ 12 , 19 – 22 ]. In this guise, we draw the idea from our own experience that mechanistic models that describe the molecular relevant parts and their qualitative interactions (such as transcription factor networks and pathways, in which the chain of causal events between genes are stated), can be enhanced in a meaningful way by adding certain porperties from the level below (e.g., physico-chemical properties that determine local interactions at the molecular component level) and by adding to the model the constraints derived from the level above (e.g., the set of emergent patterns regarding transcription factors networks from the analyses of a gene expression dataset).…”