Integrative Analysis Workflow for Untargeted Metabolomics in Translational Research

Abstract

The ability to easily quantify changes in the biological system makes metabolomics an attractive translational research tool that can help identify biomarkers of disease through non-invasive measurements Abstract Background: Metabolomics is an emerging 'omics' science that has demonstrated its fast gaining importance as a powerful profiling tool for determining an individual's response to a foreign stimulus such as a drug, toxin, or environmental change; or as an indicator of disease progression. Such small molecule profiles can used as biological markers of disease, and provide an indicator of drug efficacy or toxicity.Several studies have demonstrated that the results of any single 'omics' analysis may not be sufficient to decode extremely complex biological mechanisms. Methods:We have developed a translational research workflow to enable researchers to perform cutting-edge integrative analysis of metabolomics data with transcriptomics (gene expression) data using knowledge-driven networks. This network based view of interconnected functional partners can provide valuable new insight about the underlying biochemical processes and pathways associated with the phenotype of interest. To enhance and simplify metabolomics annotation we built a fully cross-referenced database called MetPlus DB, which integrates data from the three most comprehensive metabolite databases tailored largely towards mammalian metabolomics: HMDB, HUAMNCYC, and LIPID MAPS.Cross-referencing information is provided for linking to several other mainstream cheminformatics/bioinformatics repositories including KEGG, METLIN, ChEBI, FooDB, Pubchem, and Chemspider to provide unambiguous knowledge on clinically and physiologically relevant metabolites. We have made the integrated database available freely to the research community (https://github.com/ICBI/MetPlus-DB). Results:To demonstrate the usefulness and strength of our methodology, we have tested it on a multi-omics profiling dataset from NCI-60 breast cancer cell lines to explore the biological dynamics of breast cancer. Conclusions:Our results demonstrate a streamlined approach for the comprehensive annotation of metabolites using MetPlus DB, and the subsequent integration of metabolomics and transcriptomics data to explore potentially relevant biological interactions and candidate biomarkers associated with disease phenotype. The use of a multi-omics, systems biology approach to study cellular metabolism is key to understanding disease etiology and outcome. Integration of omics data types can help identify biological pathways affected and provides a more complete picture disease biology and prognosis, and can even provide new drug targets. One recent study integrated the metabolomics profile of tissue from breast cancer patients with protein expression data for glycerol-3-phosphate acyltransferase (GPAM), an enzyme involved in lipid biosynthesis. This data was overlaid with gene expression data to show that GPAM expression in breast cancer is associated with changes in the cellular metabol...

Metabolomics in rheumatic diseases: desperately seeking biomarkers

“…Metabolomes vary considerably between tissues and biofluids, and databases have been developed for the large amounts of information generated in studies of different systems 26 . One of the most highly utilized is the Human Metabolome Database or HMDB 27 , an open-access resource containing detailed information on >40,000 metabolites.…”

Integrative Analysis of Metabolomic, Proteomic and Genomic Data to Reveal Functional Pathways and Candidate Genes for Drip Loss in Pigs

“…The combined analysis of different omics levels provides a promising tool to increase the information density between genome and phenotype. Thereby, integrative approaches for overall analysis of the entire cascade of genome and metabolic levels (transcriptome, metabolome and proteome) provide a potential prospective to identify reliable bio markers (transcripts, proteins, metabolites) and genetic markers (SNP, QTL, candidate genes) [ 17 ]. The knowledge of functional associated omic variables/markers including interactions between genetic and environmental factors may provide a comprehensive new insight into underlying biological processes in muscle growth and meat quality [ 16 ].…”

The application of systems biology to biomanufacturing

“…A number of sophisticated statistical techniques have been developed to facilitate the analysis/integration of highly dimensional 'omics datasets including regular-ized canonical correlation analysis and sparse partial least squares regression [100]. Additional multivariate data analyses have been proposed for the integrated analyses of multiple 'omics datasets, often originating from other areas such as translational research [101][102][103][104][105]. In addition to these techniques, a more standard Pearson correlation coefficient can be computed in order to find relationships between different 'omics datasets.…”