The metabolome can be represented as a complex system characterized by a high variability of components with properties related to their aggregation and self-organizing capabilities. The elements of metabolism are molecules of various kinds and structural properties, and such variability is observed at different metabolic scales going from metabolites to metabolic profiles via chemical reactions and metabolic pathways, as well as under static or dynamic aspects. The data analysis and interpretation approaches can give insights into exciting applications of the “metabolome” studies in human health, the so-called “Metabolomics”, but the “Data Analysis” sentence should be interoperated as the global ensemble of techniques devoted to the mining of information from experimental measurements. Goals for Data Analysis applied to Metabolomics can be represented in three different classes:
Exploration of data can give answers to the questions about the structures in my data, in order to define the analytical sources of similarity and dissimiliraty properties between samples and sample aggregates, using supplemental information as meta data and covariates to correct trends of no interest in our data in order to reduce the “noise” and increase the “signal”, the information. Classification can quantify the dissimilarities/similarities between groups in order to highlight the significant changes in metabolic samples properties. Finally, Prediction can answer to the question: “What is related or predictive of my variables of interest?”
All these steps in metabolomics data analysis have received contribution from several disciplines and now we must discuss about the Computational Metabolomics as the complete technique for the physical and mathematical modeling of biomedical properties of living systems.
In order to understand the real importance of Computational Metabolomics and to understand the need to a global training of involved researchers I would like to mention in this article:
Computational Metabolomics book form Prof. Nabil Semmar and the web site of MetaboAnalyst2.0
The book presents a variety of different computational approaches to describe the variability in metabolic systems, and the Web Site of MetaboAnalyst2.0
is a comprehensive tool suite for the metabolomics data analysis. . MetaboAnalyst is supported by THE METABOLOMICS INNOVATION CENTRE (TMIC), a Genome Canada-funded core facility realized by the efforts of several Metabolomics researchers, like Jianguo Xia (our Jeff), David Wishart and Mandal, Sinelnikov, Broadhurst, Psychogios, Young et al. in the next articles I will explain to you some important aspects of the
theoretical approaches in Computational Metabolomics using the book of Prof. Semmar (and other references) and some practices with the tools of MetaboAnalyst.