Bioinformatics for spatial metabolomics

Project Information

METASPACE

Grant agreement ID: 634402

DOI

10.3030/634402

Project closed

EC signature date 27 March 2015

Start date 1 July 2015

End date 30 June 2018

Funded under

• SOCIETAL CHALLENGES - Health, demographic change and well-being

Total cost

€ 2 998 944,00

EU contribution

€ 2 998 944,00

Coordinated by

EUROPEAN MOLECULAR BIOLOGY LABORATORY
Germany

Periodic Reporting for period 2 - METASPACE (Bioinformatics for spatial metabolomics)

Reporting period: 2017-01-01 to 2018-06-30

Summary of the context and overall objectives of the project

Metabolomics is recognized as a crucial scientific domain, promising to advance our understanding of cell biology, physiology, and medicine. Metabolomics complements genomics, transcriptomics, and proteomics by analyzing the final read-out of biochemical processes and by revealing the contributions of non-genetic factors, such as the environment, diet, or microbiome. Spatial metabolomics is a next frontier, where the challenge is to localize hundreds of metabolites directly from biological tissue sections with cellular and sub-cellular spatial resolution. Our project is motivated by recent technological advances in high mass resolution imaging mass spectrometry (HR imaging MS) for imaging metabolites in biological tissue sections. Imaging MS is bringing a great promise to multiple biological and medical applications. However, a major bottleneck is still the lack of bioinformatics methods for high-throughput molecular interpretation of information-rich imaging MS data.
The overarching goal of the METASPACE project is to enable untargeted spatial metabolomics for translational research and clinical applications by providing novel bioinformatics tools, and to demonstrate their potential by using several case studies related to personalized health, precision medicine and quality of life in chronic afflictions.
The work in the project is organised by the following objectives: 1) develop novel bioinformatics for spatial metabolomics, 2) develop novel bioinformatics for knowledge-based downstream interpretation, 3) integrate state-of-the-art methods of LC-MS/MS validation into our approach, 4) create an open, accessible, user-friendly online engine for spatial metabolomics, 5) evaluate and demonstrate the online engine, raise awareness and build trust among potential users.
The consortium unites eight partners from five countries: European Molecular Biology Laboratory (international organization) participating as the headquarters EMBL Heidelberg (DE) and European Bioinformatics Institute (UK), Flanders Institute for Biotechnology (BE), Imperial College London (UK), University of California San Diego (USA), University of Rennes I (FR), SCiLS GmbH (DE), and European Research Services GmbH (DE). The partners combine expertise in metabolomics, imaging mass spectrometry, statistics, bioinformatics, and software development.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The project has developed bioinformatics for metabolite annotation of imaging MS data, published in a leading methods journal Nature Methods (Palmer et al., 2016). Using modern software development technologies and practices, we have implemented it as a high-performance cloud computing engine, deployed it onto AWS Amazon Cloud, and provided with a web app (http://metaspace2020.eu). Through providing the unique capacity in metabolite annotation supported by various dissemination efforts and trainings, we were able to engage the imaging MS community and by July 2018 have received more than 3000 submissions from over 50 labs from across the world accounting to over 100 TB of raw data. The vast majority of the annotation results are publicly shared that represents the so far largest data sharing effort in imaging MS and is actively used with over 15.000 pageviews every month.
Complementing our major efforts on metabolite annotation, we have developed a variety of bioinformatics algorithms and tools for imaging MS particularly machine learning methods for FDR-controlled annotation adapted from proteomics and for in silico fragmentation for LC-MS/MS data as well as signal processing and visualization tools for 3D molecular cartography.
We have published 20 publications including publications in high-level journals as Nature Methods, Nature Protocols, PNAS, and field-relevant journals such as Metabolomics and Analytical Chemistry. We have contributed opinions and reviews in Current Opinion in Chemical Biology and Metabolomics. We have published results of an esophageal cancer study used in our test case in one of the leading journal in cancer, Cancer Research.
The project twitter account (http://twitter.com/metaspace2020) is actively used to disseminate news and engage community and has over 390 followers. The project GitHub software repository (https://github.com/metaspace2020) hosts open-source implementations of key algorithms and software with 15 sub-repositories and over 2300 commits from ten contributors.
We have set up and keep increasing the Advisory Board which includes 25 members and serves as a key channel for dissemination of project results to academia, vendors, pharma, and journals. We organized a special session at the conferences OurCon’15 and ASMS’18 and public trainings at OurCon’16, EMSC’18 and Workshop on Imaging Mass Spectrometry’18.
The key achievements: Data for algorithm development was acquired; bioinformatics for metabolite annotation of HR imaging MS data was developed including a novel score for measuring likelihood of metabolites from a database as well as False Discovery Rate estimation approach for estimating the quality of produced annotations and selecting parameters. The scoring algorithm was improved and mapping onto KEGG metabolic pathways and genome-scale reconstructed metabolic networks were developed. The cloud software engine for metabolite annotation was developed (http://metaspace2020.eu) along with other software tools (`ili, BASIS, ChemDistiller). Proof-of-concept studies were performed by analysis of samples from cystic fibrosis, esophageal, and other cancer samples. In analysis of cancer cohorts, METASPACE has facilitated the interpretation of results not only within sample cohorts but also between cohorts. Analysing cystic fibrosis data, METASPACE helped to reveal a patient-specific metabolism of prescribed medications, differential drug penetration and microbial compartmentalisation resulting in metabolic divergence governed by local microbial interactions. These proof-of-principle studies not only helped evaluate the algorithms and software developed in the project but, importantly, successfully demonstrated how algorithms and tools developed in METASPACE can enhance data interpretation in large-scale spatial metabolomics studies.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

The novel capacity for processing thousands of datasets using the METASPACE online engine go beyond the state of the art. The annotation results accumulated from over 2800 public submissions from over 100 submitters from 50 labs represent an unprecedented knowledge base of spatial metabolomes and is already used by many users with over 15.000 page views per month. The increasing need for metabolite annotation in the imaging MS field reinforces a strong potential of the project to create a wide impact of the project onto the field of imaging MS and further in biology and medicine.
The METASPACE engine and the knowledge base are already becoming an indispensable tool in academic labs using imaging MS to address the key questions of metabolism, health and disease. We expect integration of the METASPACE engine into drug discovery and testing workflows of top pharmaceutical companies where imaging MS is rapidly gaining adoption. Moreover, we expect closer integration of the METASPACE engine into open-source software as well as imaging MS software of vendors. This will have enable answering the key questions spatial metabolomics and will have profound impact onto our understanding of metabolism in various problems of biology and medicine.

The project idea

Logo of the project