You are here

Biomex Overview


  • Biomex offers a wide set of functionalities for the analysis of metabolomics data, from raw data preprocessing to the interpretation of the results in the context of biological knowledge.
  • Biomex offers the possibility to analyze metabolic gene expression data from transcriptomics studies and integrate these results with metabolomics data.
  • Biomex is built upon the well known R statistical software and it uses a number of R/Bioconductor packages which are state-of-the-art in their respective area of application.
  • Biomex comes with a Shiny graphical interface (similar to browsing a website) and requires no programming skills.
  • Biomex results are presented as tables and figures that can be readily included in scientific publications.
  • Biomex is open-source and freely available to the academic metabolomics community. 
Liquid chromatography mass spectrometry (LCMS)-based untargeted metabolomics is a validated technology that is increasingly applied in diverse fields ranging from analytical chemistry to clinical medicine. The ultimate goal of untargeted metabolomics is to provide a comprehensive metabolic profile from which new biological insight can be derived. However, the translation of high-density metabolic profiling data into concrete biological hypotheses remains challenging. Technical limitations such as metabolite identification, instrument sensitivity and robustness  have to be addressed but it is becoming increasingly clear that, on the long-term, the biggest challenge for metabolic profiling is not of technological nature but deriving actionable knowledge from the vast amounts of data produced by these technologies.

 A plethora of very useful and popular bioinformatics tools are now available to improve biological interpretation of metabolomics experiments. However, these tools often require advanced bioinformatics skills and are focused on a single task in the metabolomics data analysis workflow. The value of individual tools could be maximized when integrated in a unified user-friendly multi-omics metabolic analysis workflow that is focused on experimental scientists.

To this end we, together with an international team of collaborators, are developing an integrated R and Shiny-based open-source software that aims to facilitate Biological Interpretation Of Metabolomics EXperiments (BIOMEX). With BIOMEX we aim to merge publically available algorithms with in-house developed software to create a unified workflow for integrated analysis of MS-based metabolomics and transcriptomics (metabolic gene expression data). BIOMEX will perform raw data preprocessing, identify metabolites against a user-defined database, perform untargeted tracer fate detection in stable-isotope assisted experiments and provide functionality for univariate and multivariate statistical analysis, including multigroup experiments. Results of statistical analysis can be contextualized through integrative metabolite and gene set enrichment analysis and pathway mapping. BIOMEX allows the user to interactively select relevant features at all steps of the workflow and search the HMDB metabolite and NCBI GENE databases to obtain detailed information. Users may download a dynamic HTML report for easy distribution of analyzed experiments and/or download the results in a format that is easily uploaded to the metabolights database. While we are still developing many aspects of BIOMEX, certain modules are in advanced stages - manually checked and validated results have been incorporated in a number of publications.

The image below (adapted from Rolin 2012) depicts the BIOMEX workflow. Briefly, users may input raw LCMS metabolomics data (including samples from stable isotope-assisted experiments), preprocessed transcriptomics data (raw counts from RNA-sequencing or  intensities from microarray experiments) or search and upload data from a gene expression repository. Metabolomics, tracer and transcriptomics experiments may be processed at the same time and all experiments can have multiple groups. In addition, it is possible to batch analyze multiple experiments. BIOMEX provides functions for data quality control including metabolomics preprocessing (peak integration, peak shape, grouping, etc). Users can upload a metabolite database for metabolite identification or search the HMDB library for specific features (considering common adducts). After data preprocessing and quality checks, BIOMEX provides functions for flexible data analysis, the design of experiment and thus research question can be updated on the fly and answered using univariate, multivariate and bioinformatics analysis. The user may provide multiple variables (factors, covariates, group assignment, etc) which will be included in the analysis using linear and mixed models (in case random factors are included). The results of statistical, bioinformatics and tracer analysis can be forwarded to meta-analysis or used as input for genome-scale metabolic modeling. Thus BIOMEX aims to provide an easy to use interface to perform flexible but detailed data analysis.



Karen Vousden, Paolo Sassone-Corsi, Christian Frezza, Nika Danial
12/09/2017 - 09:00