Metabolomic Modularity Analysis (MMA) to Quantify Human Liver Perfusion Dynamics.
Gautham Vivek Sridharan, Bote Gosse Bruinsma, Shyam Sundhar Bale, Anandh Swaminathan, Nima Saeidi, Martin L Yarmush, Korkut Uygun
Author Information
Gautham Vivek Sridharan: Center for Engineering in Medicine, Harvard Medical School, Massachusetts General Hospital & Shriners Hospital for Children, 51 Blossom Street, Boston, MA 02114, USA. gvsridharan@gmail.com.
Bote Gosse Bruinsma: Center for Engineering in Medicine, Harvard Medical School, Massachusetts General Hospital & Shriners Hospital for Children, 51 Blossom Street, Boston, MA 02114, USA. botebruinsma@gmail.com.
Shyam Sundhar Bale: Center for Engineering in Medicine, Harvard Medical School, Massachusetts General Hospital & Shriners Hospital for Children, 51 Blossom Street, Boston, MA 02114, USA. shyam.bale@gmail.com.
Anandh Swaminathan: Department of Control and Dynamic Systems, California Institute of Technology, Pasadena, CA 91125, USA. aswamina@caltech.edu.
Nima Saeidi: Center for Engineering in Medicine, Harvard Medical School, Massachusetts General Hospital & Shriners Hospital for Children, 51 Blossom Street, Boston, MA 02114, USA. Nsaeidi@mgh.harvard.edu.
Martin L Yarmush: Center for Engineering in Medicine, Harvard Medical School, Massachusetts General Hospital & Shriners Hospital for Children, 51 Blossom Street, Boston, MA 02114, USA. ireis@sbi.org.
Korkut Uygun: Center for Engineering in Medicine, Harvard Medical School, Massachusetts General Hospital & Shriners Hospital for Children, 51 Blossom Street, Boston, MA 02114, USA. Uygun.Korkut@mgh.harvard.edu.
Large-scale -omics data are now ubiquitously utilized to capture and interpret global responses to perturbations in biological systems, such as the impact of disease states on cells, tissues, and whole organs. Metabolomics data, in particular, are difficult to interpret for providing physiological insight because predefined biochemical pathways used for analysis are inherently biased and fail to capture more complex network interactions that span multiple canonical pathways. In this study, we introduce a nov-el approach coined Metabolomic Modularity Analysis (MMA) as a graph-based algorithm to systematically identify metabolic modules of reactions enriched with metabolites flagged to be statistically significant. A defining feature of the algorithm is its ability to determine modularity that highlights interactions between reactions mediated by the production and consumption of cofactors and other hub metabolites. As a case study, we evaluated the metabolic dynamics of discarded human livers using time-course metabolomics data and MMA to identify modules that explain the observed physiological changes leading to liver recovery during subnormothermic machine perfusion (SNMP). MMA was performed on a large scale liver-specific human metabolic network that was weighted based on metabolomics data and identified cofactor-mediated modules that would not have been discovered by traditional metabolic pathway analyses.
