A comprehensive online spectra analysis and visualisation tool for the OMICS sciences

Metabolomics is a highly data-generating and knowledge-driven science. The study of the small molecule complement of an organism creates large amounts of information-rich data that provide unprecedented insights into an organism's biology. In metabolomics, evidence of the occurrence of small molecules as well as their concentrations and role in signalling and regulatory pathways is acquired within different biological levels such as the organism, tissue, cell type, or compartment level. Analysis and interpretation of metabolomics data - through utilisation of chemical and biological databases - brings together information from different resources to explain chemical identities, structures, and biological function.

Metabolomics resources contain all evidence from a study such as metabolite structures and their reference spectra, biological roles, locations and concentrations, as well as experimental data. Comprehensible visualisation of metabolomics data is extremely difficult because of the plethora of information that is inherently connected to it. There is high demand for a comprehensive web-based visualisation and analysis tool for metabolomics to enable efficient data exploration and visualisation.

Existing databases for metabolomics store information about identified chemical structures and their spectroscopic fingerprints, partially with additional annotations. The databases are designed to aid in compound identification through simple queries. The data visualisation capability of these databases is thus limited to specific tasks. Depending on the database, visualisation ranges from static images of spectra to interactive solutions that show spectra including signal annotations. The interactive solutions allow some control over the visualisation but lack functionality to display textual and pictorial annotations beyond the bare minimum. These viewers have been developed as part of individual databases and are deeply integrated into their systems.

Few existing online spectrum viewers are good for visualising results with some degree of spectral comparison capabilities or database queries, and only show partial information. It is time consuming and challenging to develop a generic online spectrum viewer with advanced charting functionality that supports different views and platforms (NMR, MS, and chromatography). Such a tool is crucial to enable metabolomics resources to be developed and used to their full potential and to be interactive. With no flexible full-fledged viewer currently available, we will develop a comprehensive online spectrum viewer for metabolomics that is enriched in such functionality. We will demonstrate practical usage of the tool in the BBSRC-funded cross-species, cross-platform MetaboLights database and repository for metabolomics data. We are committed to make the tool open-source, freely available, and to use permissive licensing to encourage collaboration and continuous development.

Mass Spectrometry and Nuclear Magnetic Resonance are the two principal methods used in metabolomics experiments. Analysed data sets usually contain textual and symbolic annotations, as well as cross-references to database entries. The big challenge is to comprehensively visualise these data sets and associated information in an intuitive, interactive, and intelligent manner that adheres to established metabolomics standards.

Currently there is no versatile online spectrum viewer available for metabolomics that can be conveniently adapted and integrated with such contextual information. A dedicated open-source solution will concentrate efforts currently spent on patchwork and usually incomplete solutions. A full-fledged web-based visualisation tool will benefit the metabolomics community by giving an unparalleled view of available data.

The technology of choice is JavaScript. It is one of the most popular programming languages on the web. An invaluable advantage of JavaScript is that it can be used in a way that it runs on any modern browser without the need for installing additional plugins or software. The rise of AJAX has furthermore created toolkits that allow for the authoring of JavaScript-based web applications in a browser-independent way.

To ensure easy integration into existing metabolomics databases, we will create an intuitive and well-documented application programming interface (API). Together with the benefit of lacking external dependencies and cross-browser compatibility, the viewer will be easy to adapt in metabolomics databases. Furthermore, we will provide convenience methods to generate views such as 1D and 2D NMR spectra or difference spectra. Database links for the dominant chemistry and metabolomics resources will be established to assist with resolving external identifiers. Integration into MetaboLights will require common input functionality and usage of open metabolomics file standards, such as mzML.

The proposed visualisation tool will benefit the scientific community, industry, and general public by driving metabolomics research. Metabolomics is key to fully understand biological processes and its non-invasive nature and extremely high time resolution designate it as an ideal tool for the pharmaceutical industry. Biomarker discovery and drug safety screens are two examples where metabolomics has enabled informed decision making. Samples of urine, saliva, bile, or seminal fluid contain the necessary information and can be readily analysed through metabolomics fingerprinting or profiling. Drug development, translational research, and clinical trials - from the preclinical phase to in vivo testing on patients - also benefit heavily from metabolomics due to the high correlation between the metabolome and the phenotype. General healthcare and personalised medicine use metabolomics in medical diagnostics to quickly establish health profiles and discover disease states. A full-featured spectrum viewer will facilitate informative data exploration and drive further research in metabolomics. This will also enable a feedback loop into the tool's development as it will be more widely used and once more features were added.

Food safety and clinical microbiology (monitoring) draw from the reproducibility of metabolomics fingerprints that allow reliable identification of contaminants, helping to improve production standards and hygiene. Rapid identification of potentially harmful strains of bacteria is essential in modern hospitals to allow medical staff to make informed decisions.

The tool will be published in an open access journal and its usage tracked using internet. Access statistics and download statistics will be set up for the full-featured online version and source repository respectively.
Within the context of the European Bioinformatics Institute, a highly successful Bioinformatics Training programme is provided by a dedicated Outreach and Training team. This training programme is provided to increase adoption and impact of the EBI resources, and will be harnessed to ensure maximal impact through training for adoption and use of the proposed tool for all industrial and public sector beneficiaries. Furthermore, the EBI runs a successful Industry Programme where targeted workshops and training are provided to members of participating industrial organisations, increasing the impact of EBI resources including our BBSRC-funded Metabolomics database and the tool which we are here proposing to develop.