- Genomics research
Metabolomics has been used for obtaining direct measures of metabolic activities from diverse biological systems. However, metabolomics can be limited by ambiguous metabolite identifications. Furthermore, interpretation can be limited by incomplete and inaccurate genome-based predictions of enzyme activities (e.g., gene annotations). In addition, some genes may be poorly annotated. Thus, the understanding of metabolism, such as microbial metabolism, is limited.
A team of Lawrence Berkeley National Laboratory researchers including Trent Northen, Benjamin Bowen, Oliver Reubel, and Markus De Raad have developed technologies for associating metabolites with genes that enable scoring and curating compound identities based on their biological relevance, and/or using compound identities from those tools to connect to genes in their biological samples and potentially formulate hypotheses of gene function. Such results can be used to direct high-throughput biochemical assays to greatly reduce biochemical search space. The Metabolite, Annotation and Gene Integration (MAGI) system is highly relevant to and useful in the fields of genomics, metabolomics, and systems biology. Furthermore, as metabolomics data become more widely available for sequenced organisms, MAGI has the potential to improve the understanding of microbial metabolism, while also providing testable hypotheses for specific biochemical functions.
MAGI v.1.0, a computational software package developed by researchers Onur Erbilgin, Benjamin Bowen, and Oliver Ruebel, efficiently supports the above workflow, integrating experimental metabolomics and genomics data with chemical, biochemical, and genomic data to produce and test hypotheses. In the MAGI workflow, exact chemicals are linked to exact genes via probabilistic relationships between reactions facilitated by a chemical similarity network, and protein homology searching. MAGI automatically suggests alternative substrates to experimentally test via the chemical similarity network, necessary to determine the specific function of an enzyme, and can execute the experiments in a high throughput manner. MAGI is useful for compound identifications in untargeted metabolomics experiments and annotating genes and genomes, and is most powerful when the two are combined. It is also a major aide for biochemical function discovery, biosynthetic pathway (re)construction, metabolic modeling, and many more aspects of biochemistry. This overcomes previous approaches where researchers need to use several disconnected resources to connect a gene to a compound in a reaction.
STATUS: Published U. S. Patent Application 15/932,459 (Publication No. US2018-0239863). Software copyrighted. Available for licensing or collaborative research.
REFERENCE: 2017-063, 2017-105