Top image

Contact us

For any enquiries related to the Computational Medicine research area, please contact

Maximising information recovery from complex biological data

Researchers in CSM develop and deliver novel tools for the analysis, integration, and visualisation of complex biological datasets that maximise the benefit that can be derived from top-down systems biology approaches to understanding human health and disease. The Division have been responsible for a number of recent developments in the field of post-genomic dataset analysis, including the development of expert systems for toxicity classification, application of genetic algorithms to optimise sample and variable selection, spectral alignment and calibration.  

Combining information from multiple omics platforms promises to allow a more holistic approach to understanding the complexity of mammalian biochemistry and CSM continue to develop novel statistical and pathway-based approaches for data integration.  Recovering the latent information related to subtle time-modulated changes present in complex overlapped spectra is a considerable challenge. We have developed curve resolution methods based on the concept of spectrotype subunits of global profiles that enable such subtle features to be efficiently revealed.

Find out more about our key focus areas:

Accordion widget

Statistical Methods for Dataset Linkage and Integration

The development of statistical spectroscopy methods for information recovery from complex biofluid datasets has been driven forward in recent years and is now established as an efficient way of interrogating spectral data. Used independently, NMR and MS are useful for structure elucidation, and can provide much of the necessary information required for identifying compounds through the use of various experiments (e.g. the use of multidimensional NMR experiments to indicate bonding of different functional groups; use of tandem MS experiments to generate parent and daughter ion data). While experienced analysts may be able to integrate such data, we have developed statistical linkage of spectral sets to aid this process that make use of the (typically) large number of samples in metabolic profile studies; these methods utilise the relative  the strength of correlation between spectral features across many samples to indicate those most likely to have arisen from the same metabolite. These methods may also help delineate sets of metabolite responding in a concerted manner.

CSM have participated in collaborative efforts to provide metabolic pathway/network based tools for integrating metabolic phenotype data with other biological measures; to enhance our understanding in the genetic regulation of complex metabolic patterns in a network biology framework, candidate genes and metabolic biomarkers are mapped onto biological networks such as metabolic pathways or protein-protein interaction networks to identify key regulatory proteins explaining the influence of gene variants on metabolic profiles and eventually disease phenotypes. Recent developments such as Metabolite-Set Enrichment Analysis (MSEA) or integrated Metabolome Interactome Mapping (iMIM) allow a better mechanistic understanding of metabolic phenotypes.  

Reporting Standards and Data Repositories

The chemical space represented by cellular, organ and systemic metabolism is vast, and a large number of different analytical approaches have been developed to capture facets of the whole; no one method captures adequately information on all metabolome components as a result of inherent sensitivity and selectivity limitations. One consequence is that deriving standard data formats and reporting is complex.  Members of CSM are involved in community efforts to improve the quality literature reporting on the confidence of spectral assignments in metabonomics studies, particularly in relation to structure elucidation of previous unknowns, to prevent propagation of assignment errors. 

Quantitative Genetics and Network Biology of Metabolic Profiles

The study of human multifactorial diseases like cancer, diabetes, obesity or cardiovascular disease and complex biological processes such as ageing, represents a real healthcare challenge for the western and developing world. Regulation of metabolism and signalling is a key biological phenomenon in these conditions and the field of metabonomics has now made a significant impact in functional genomics, owing to serendipitous biomarker discovery in model organisms and in human populations, as well as candidate-driven approaches. In large-scale epidemiological studies (n>1,000), association between metabolic phenotypes and disease phenotypes gave rise to Metabolome-Wide Association Studies (MWAS).

However, the study of the genetics of metabolic profiles is a wide area, ranging from characterizing genetically-engineered model organisms or inborn errors of metabolism to the quantitative study of metabolic phenotypes in segregating populations. Quantitative Trait Locus (QTL) mapping of metabolomic traits (mQTL), and Metabolome-Wide Genome-Wide Association Studies (MW-GWAS) are robust and accurate strategies for the integration of genome-wide genotyping and metabolome-wide profiling by 1H NMR and MS, identifying candidate biomarkers and susceptibility genes. 

Finally, to enhance our understanding in the genetic regulation of such complex metabolic patterns in a network biology framework, candidate genes and metabolic biomarkers are mapped onto biological networks such as metabolic pathways or protein-protein interaction networks to identify key regulatory proteins explaining the influence of gene variants on metabolic profiles and eventually disease phenotypes. Recent developments such as Metabolite-Set Enrichment Analysis (MSEA) or integrated Metabolome Interactome Mapping (iMIM) allow a better mechanistic understanding of metabolic phenotypes. 

Key members within Computational Medicine

Search or filter publications

Filter by type:

Filter by publication type

Filter by year:



  • Showing results for:
  • Reset all filters

Search results

This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.

Request URL: Request URI: /respub/WEB-INF/jsp/search-t4-html.jsp Query String: id=923&limit=5&respub-action=search.html Current Millis: 1558631845262 Current Time: Thu May 23 18:17:25 BST 2019