1. Overview

Metabolomics is a rapidly developing field of -omics research concerned with the high-throughput identification and quantification of small molecule metabolites in the metabolome. The metabolome constitutes a wide array of compound classes that are crucial for the normal functioning of a biological system. As a result, the metabolomics approach promises to offer new insights in many areas of biological investigation. 

Recent metabolomics research benefited greatly from advances in mass spectrometry and chromatography. These advances allow researchers to detect many metabolites that could not be detected previously. However, a sizable fraction of these compounds is unknown, and a new computational infrastructure is required for processing the complex mass spectral data and facilitating the identification of these metabolites. 

The PI's group addressed this need through this awarded project by having developed a streamlined computational workflow that can create libraries of high-quality mass spectra for unknown compounds from many biological samples. Furthermore, the PI's group has used this computational workflow and created such a mass spectral library for facilitating prioritization and eventual identification of currently unidentified metabolites. Equally importantly, the developed workflow allows more precise quantitation of metabolites and subsequent differential analysis of metabolic profiles. The most biologically interesting unknown compounds in the library can then be subjected to further attempts at structure elucidation.

2. List of grant awards, publications and presentations

With the results from this project, the PI applied for NIH funding and has been awarded two grants from the NIH Common Fund Metabolomics Program with a combined total cost at $1.53 million. In addition, the PI's group has published five journal papers and given about 35 conference/workshop oral presentations and five conference/symposium poster presentations. One journal manuscript is currently near completion and another one is in preparation. NSF funding was acknowledged in all of the publications and presentations.

3. Intellectual Merit

The computational workflow that resulted from this project streamlined the process of extracting reliable mass spectra of single compounds from complex raw mass spectral data and creating libraries of unknown spectra. It removed obstacles to exhaustive metabolite profiling, allowed researchers to build spectral libraries representing unknown compounds of potential interest, and paved the way for subsequent characterization and structure elucidation of these unknowns. Eventually, knowledge gained about the unknown compounds enables researchers to: 1) synthesize them, 2) discover new metabolic pathways, 3) conduct further studies to understand the roles of these metabolites in metabolic pathways, 4) discover interactions between the metabolome and other -omes (i.e., genome, transcriptome, and proteome), 5) yield valuable insights into underlying biochemical processes, and 6) ultimately contribute to advances in functional genomics. The computational workflow could especially benefit plant science and related areas because the plant kingdom is estimated to have more than 200,000 metabolites of enormous biochemical diversity, and many of them are unknown. 

4. Broader Impacts

The project has enabled the PI to train next-generation bioinformaticians by training two postdoctoral researchers, one Ph.D. student, and four master's students. In addition, the PI has been invited to speak for six consecutive years (2013-2018) at the UAB metabolomics workshop, which was sponsored by the National Institute of General Medical Sciences (NIGMS) as part of the NIH Common Fund Metabolomics Initiative. Each year, a total of about 40 participants attended the workshop. These included graduate students, basic science and clinical fellows, junior and mid-career faculty members, research associates, and directors of facilities. These six workshops trained a large number of scientists who could start using metabolomics in their own research. Last but not the least, the PI has used metabolomics examples and datasets accumulated over the course of the project and used them in teaching undergraduate and graduate courses including Machine Learning for Bioinformatics in Fall 2014, Fall 2015, Fall 2016, Fall 2017, Statistics for Bioinformatics in Spring 2015, Spring 2016, and BINF 4211: Applied Data Mining for Bioinformatics in Spring 2018. 

Last Modified: 01/08/2019
Modified by: Xiuxia Du