Abstract
The functional connection of experimental metabolic time series data with biochemical network information is an important, yet complex, issue in systems biology. Frequently, experimental analysis of diurnal, circadian, or developmental dynamics of metabolism results in a comprehensive and multidimensional data matrix comprising information about metabolite concentrations, protein levels, and/or enzyme activities. While, irrespective of the type of organism, the experimental high-throughput analysis of the transcriptome, proteome, and metabolome has become a common part of many systems biological studies, functional data integration in a biochemical and physiological context is still challenging. Here, an approach is presented which addresses the functional connection of experimental time series data with biochemical network information which can be inferred, for example, from a metabolic network reconstruction. Based on a time-continuous and variance-weighted regression analysis of experimental data, metabolic functions, i.e., first-order derivatives of metabolite concentrations, were related to time-dependent changes in other biochemically relevant metabolic functions, i.e., second-order derivatives of metabolite concentrations. This finally revealed time points of perturbed dependencies in metabolic functions indicating a modified biochemical interaction. The approach was validated using previously published experimental data on a diurnal time course of metabolite levels, enzyme activities, and metabolic flux simulations. To support and ease the presented approach of functional time series analysis, a graphical user interface including a test data set and a manual is provided which can be run within the numerical software environment Matlab®.
Highlights
The functional interpretation of experimental data in context of biochemical network information represents one of the central challenges in current biological research
Time-dependent dynamics of metabolite concentrations in a biochemical network can be described by a set of ordinary differential equations (ODEs): d dt
M represents an n-dimensional vector of mean metabolite concentrations, N is the n × k stoichiometric matrix and v describes the k-dimensional vector of reaction rates which depend on metabolite concentrations M, enzyme kinetic parameters p and time t
Summary
The functional interpretation of experimental data in context of biochemical network information represents one of the central challenges in current biological research. Genome sequence information and metabolic networks have become available for Metabolomic Time Series Analysis numerous organisms, tissues, or cell types (Herrgard et al, 2008; Chang et al, 2011; De Oliveira Dal’Molin and Nielsen, 2013; Thiele et al, 2013), functional metabolic data interpretation still represents a major obstacle in systems biology. Various mathematical and computational strategies from the fields of multivariate statistics, ordinary, and partial differential equations (ODEs/PDEs), optimization or statistical time series analysis have been developed and applied to reveal a biologically meaningful interpretation of comprehensive and multidimensional experimental data sets. The authors developed a model of 28 ODEs which were numerically simulated in order to analyze diurnal kinetics of carbon metabolism in silico. The authors derived a prediction about the impact of product formation on biomass concentration using steady state simulations at varying environmental conditions
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
