The long-term objective of the research described in this application is to develop computational, analytical and database tools that can be used to rapidly identify the chemical structure of compounds in human biofluids. These computational and analytical tools will significantly improve the use of metabolomics for: a) understanding disease mechanisms, b) allowing earlier disease diagnosis, and, c) enhancing the accuracy of disease prognosis. Our innovative approach is to develop algorithms that predict physical/chemical properties of compounds contained in PubChem and other large databases. The physical/chemical properties chosen are those that can be experimentally measured for any unknown compound by HPLC-mass spectrometry. Compounds in chemical databases whose predicted properties most closely match experimental properties are returned as the most likely candidates for the unknown. We will then apply these tools to an animal model of hemorrhagic shock and in children with acute trauma. Our previous work and preliminary data describe the validity of this approach using computational models developed for predicting retention indices, Ecom50, drift index and collision induced dissociation fragmentation spectra. Based on these promising preliminary data, we propose the following specific aims:
Specific Aim 1 : Optimize current and develop new predictive models (positive ion Ecom50 and positive ion drift index) to allow efficient searching of large databases using MolFind/BioSM.
Specific Aim 2 : Build a virtual biochemical database of >2x106 compounds by generating in silico human phase I and phase II metabolites of all compounds in the KEGG, HMDB, FooDB, DrugBank, HumanCyc, Metlin, PlantCyc, Phenol Explorer and Lipid Maps databases.
Specific Aim 3 : Optimize input parameters to develop a database assisted structure identification (DASI) tool that will produce rationally designed virtual candidate metabolites.
Specific Aim 4 : Validate the use of MolFind/BioSM for non-targeted metabolomics in a swine model of hemorrhagic shock and for hypothesis driven targeted metabolomics in children with acute trauma. By facilitating the rapid structural identification of chemical compounds in clinically relevant biofluids, these innovative tools will greatly enhance the ability of metabolomics studies to compliment and synergize other areas of biomedical research and ultimately improve human health care.

Public Health Relevance

Metabolite levels in cells, tissues and biofluids provide a wealth of information related to organism function, thus allowing a unique understanding of disease mechanisms. This application is designed to provide the computational tools and analytical methods necessary to access and interpret this information in order to improve human health in general and the clinical management of children with acute trauma in particular.

National Institute of Health (NIH)
National Institute of General Medical Sciences (NIGMS)
Research Project (R01)
Project #
Application #
Study Section
Biodata Management and Analysis Study Section (BDMA)
Program Officer
Sheeley, Douglas
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Connecticut
Schools of Pharmacy
United States
Zip Code
Menikarachchi, Lochana C; Dubey, Ritvik; Hill, Dennis W et al. (2016) Development of Database Assisted Structure Identification (DASI) Methods for Nontargeted Metabolomics. Metabolites 6:
Paglia, Giuseppe; Angel, Peggi; Williams, Jonathan P et al. (2015) Ion mobility-derived collision cross section as an additional measure for lipid fingerprinting and identification. Anal Chem 87:1137-44
Hamdalla, Mai A; Rajasekaran, Sanguthevar; Grant, David F et al. (2015) Metabolic pathway predictions for metabolomics: a molecular structure matching approach. J Chem Inf Model 55:709-18
Dubey, Ritvik; Hill, Dennis W; Lai, Steven et al. (2015) Correction of precursor and product ion relative abundances in order to standardize CID spectra and improve Ecom50 accuracy for non-targeted metabolomics. Metabolomics 11:753-763
Hall, L Mark; Hill, Dennis W; Menikarachchi, Lochana C et al. (2015) Optimizing artificial neural network models for metabolomics and systems biology: an example using HPLC retention index data. Bioanalysis 7:939-55
Paglia, Giuseppe; Williams, Jonathan P; Menikarachchi, Lochana et al. (2014) Ion mobility derived collision cross sections to support metabolomics applications. Anal Chem 86:3985-93
Hall, L Mark; Hill, Dennis W; Hall, Lowell H et al. (2014) Development of HPLC Retention Index QSAR Models for Nontargeted Metabolomics. Adv Chromatogr 51:241-79
Hall, Lowell H; Hall, L Mark; Hill, Dennis W et al. (2014) Development of a two-step indirect method for modeling Ecom50. Curr Comput Aided Drug Des 10:374-82
Hamdalla, Mai A; Mandoiu, Ion I; Hill, Dennis W et al. (2013) BioSM: metabolomics tool for identifying endogenous mammalian biochemical structures in chemical structure space. J Chem Inf Model 53:601-12
Menikarachchi, Lochana C; Hamdalla, Mai A; Hill, Dennis W et al. (2013) Chemical structure identification in metabolomics: computational modeling of experimental features. Comput Struct Biotechnol J 5:e201302005

Showing the most recent 10 out of 14 publications