MIRA: mutual information-based reporter algorithm for metabolic networks

Abstract

Motivation: Discovering the transcriptional regulatory architecture of the metabolism has been an important topic to understand the implications of transcriptional fluctuations on metabolism. The reporter algorithm (RA) was proposed to determine the hot spots in metabolic networks, around which transcriptional regulation is focused owing to a disease or a genetic perturbation. Using a z-score-based scoring scheme, RA calculates the average statistical change in the expression levels of genes that are neighbors to a target metabolite in the metabolic network. The RA approach has been used in numerous studies to analyze cellular responses to the downstream genetic changes. In this article, we propose a mutual information-based multivariate reporter algorithm (MIRA) with the goal of eliminating the following problems in detecting reporter metabolites: (i) conventional statistical methods suffer from small sample sizes, (ii) as z-score ranges from minus to plus infinity, calculating average scores can lead to canceling out opposite effects and (iii) analyzing genes one by one, then aggregating results can lead to information loss. MIRA is a multivariate and combinatorial algorithm that calculates the aggregate transcriptional response around a metabolite using mutual information. We show that MIRA’s results are biologically sound, empirically significant and more reliable than RA.Results: We apply MIRA to gene expression analysis of six knockout strains of Escherichia coli and show that MIRA captures the underlying metabolic dynamics of the switch from aerobic to anaerobic respiration. We also apply MIRA to an Autism Spectrum Disorder gene expression dataset. Results indicate that MIRA reports metabolites that highly overlap with recently found metabolic biomarkers in the autism literature. Overall, MIRA is a promising algorithm for detecting metabolic drug targets and understanding the relation between gene expression and metabolic activity.Availability and implementation: The code is implemented in C# language using .NET framework. Project is available upon request.Contact: cicek@cs.cmu.eduSupplementary information: Supplementary data are available at Bioinformatics online

A Conversation on Data Mining Strategies in LC-MS Untargeted Metabolomics: Pre-Processing and Pre-Treatment Steps

“…[12,43], call for caution in handling noise elimination and redundant signals in metabolomic datasets. The effort to produce a noise-free and non-redundant data matrix (post-processing) could result in a loss of information, some of which could be informative to comprehensively assess metabolic pathways for a better understanding and description of the regulatory mechanisms underlying the global biological responses [12,18,37,44,45]. Furthermore, as recently demonstrated, some of the ion peaks that could be regarded as a source of redundancy (e.g., adduct formation) might be very crucial and actually needed in metabolite annotation and differentiation [46].…”

“…In order to maximize the value of metabolomic data and generate biologically-meaningful hypotheses, particularly with regard to the regulatory mechanisms and molecular processes involved in global biological responses (such as those of a biosystem), the metabolomics raw data are to be appropriately handled and fully exploited. This will ensure the extraction of sufficient information to determine, as holistically as possible, biological components that show differential behaviors between experimental conditions [1,11,12,18,19]. …”

Reporter pathway analysis from transcriptome data: Metabolite-centric versus Reaction-centric approach

“…In the metabolite-centric reporter pathway analysis (RPA m ), a metabolite-score must be computed first. This scoring method, termed reporter metabolite analysis, was used in a number of research covering microorganisms 7 52 and health problems such as liver diseases, obesity, autism 53 54 55 . Different versions of the approach also appeared 48 56 .…”

Updates in metabolomics tools and resources: 2014–2015

“…In a very innovative approach to capture the micro‐organismal, interspecies competition‐induced, highly complex, ecological interactions‐generated, untargeted metabolomics data of bioactives in “cocultures,” Projected Orthogonal CHemical Encounter MONitoring (POCHEMON), written in MATLAB, was shown to be a novel multivariate data analysis method that reveals all competition‐related, biochemical changes from the cocultures, both up‐ or downregulated, and de novo synthesized metabolites . Similarly, mutual information‐based multivariate reporter algorithm (MIRA) is a multivariate and combinatorial algorithm that calculates the aggregate transcriptional response around a metabolite using mutual information . In this proof‐of‐concept study, MIRA was successful in the gene expression analysis of six knockout strains of Escherichia coli to help capture the metabolic dynamics of the switch from aerobic to anaerobic respiration by eradicating statistics related issues associated with small sample sizes, analyses of genes one‐by‐one and as a group, and overcoming z‐score associated inconsistencies.…”

“…Genome‐scale metabolic networks have caught the attention of biological researchers, as they assist in adding insights into biological questions, as compared to the much more simplistic view of one‐dimensional biochemical pathways such as KEGG. PathCase Metabolomics Analysis Workbench (PathCaseMAW) runs on a manually created, generic, mammalian metabolic network, is a database‐enabled framework (with its user‐friendly interface), and can be used to generate new metabolic networks and/or update an existing metabolic network . These genome‐scale metabolic networks are accessible through a web interface or an iPad application that implement an integrated steady‐state metabolic network dynamics analysis algorithm.…”