A recent research paper published in Nature Communications proposes an innovative workflow to study engineered microbes in synthetic biology applications. The researchers, from the Pacific Northwest National Laboratory, have combined advanced analytical instrumentation with a machine learning-based algorithm to better identify metabolites, the small molecules produced by cellular processes that can provide information on the health and activity of microorganisms.
This workflow makes use of a combination of Liquid Chromatography, Ion Mobility and Data-Independent Acquisition Mass Spectrometry, allowing for the rapid analysis of multiple samples and the extraction of valuable metabolite information. In order to process the complex data generated, Bilbao has developed a novel PeakDecoder algorithm that learns to differentiate between different molecule information from the raw data. This algorithm provides faster analysis, with a third of the sample processing time of traditional older approaches.
To illustrate the capabilities of the workflow, it was used to study engineered microorganisms for the production of different bioproducts. Using complex mixture measurements, 2,683 metabolite features from 116 samples were correctly identified across various microbial strains.
The research team believes that the machine learning-based software could replace traditional methods that require large-scale human intervention with more sophisticated artificial intelligence-based solutions. The developers aim to create a user-friendly, fully automated version of PeakDecoder over the coming months, which should allow it to support research projects on a large scale.
Technology Networks Ltd is a global platform providing current scientific news, commentary and events online. They provide resources to follow research progress, allowing members of the scientific community to stay up-to-date on the latest industry developments. They offer products and services to help organizations and researchers to monitor, measure and compare within the industry.
Dr. Kristin Burnum-Johnson is a biochemist and lead at the Agile BioFoundry Task. The Agile BioFoundry is a platform supported by the US Department of Energy that provides a program for the development of bioproducts using advanced engineering for synthetic biology applications.
Dr. Aivett Bilbao is a computational scientist at the Pacific Northwest National Laboratory. He is the developer of the PeakDecoder algorithm, which can identifiy molecules in complex mixtures and accurately profile multidimensional mass spectrometry data from large scale studies. He is currently working on the next version of the algorithm, which should improve the performance and automate the whole process.
