Virginia Commonwealth University

This figure is a simulation of LC-DAD chromatograms of an evolving chemical system.

Rutan Home



Research Group

Research Projects








Chemometrics and Comprehensive Two-Dimensional Liquid Chromatography:
Toward Achieving the Promise for Metabolomics

Research in the Rutan group focuses on understanding the fundamental principles that underlie the quantitative capabilities of two-dimensional liquid chromatography (2D-LC) and developing methods for improvement of the performance of quantitative 2D-LC methods. This work is targeted towards enhancement in the use of this methodology for the study of complex biological systems, especially in the area of metabolomics. Specifically, this work addresses the optimization of the separation conditions, the exploration of novel means to accelerate the 2nd dimension separation with an emphasis on the examination of the resulting data by novel chemometric methods designed to overcome the challenges currently associated with 2D-LC-DAD analysis of complex samples. The samples of interest in plant and animal metabolomic studies typically contain 100's to 1000's of compounds. One component of the work is the development of optimization strategies to obtain fast and effective 2D-LC separations of these complex mixtures. Methods employed include the use of the Snyder-Dolan "hydrophobic subtraction" method to find maximally orthogonal reversed-phase columns for the 1st and 2nd dimension separations and the assessment of the performance of the resulting methods via a multivariate selectivity metric. A major focus of the work is in the development of the essential chemometric methods necessary for the automated transformation of the raw signals into precisely defined (5-10% RSD) chemical compositions of large numbers of samples whose chromatograms show 100's of peaks. This part of the work capitalizes on novel developments in multi-way data analysis methods including multivariate curve resolution-alternating least squares (MCR-ALS) and parallel factor analysis (PARAFAC) methods. The use of a diode array detector (DAD) is most suited for this purpose as it is very robust and not sensitive to serious matrix effects. The goal of the work are to develop protocols that include methods for determining the number of components present in the data, correcting for misalignment of the time base of the chromatograms due to retention time drift, resolving overlapped chromatographic peaks, and precisely quantifying the relative concentration of compounds present in complex mixtures. This project includes developing a fundamental understanding of how changes in the chemistry of the chromatographic process affect the characteristics of the resulting data, understanding how these chemical factors affect the performance of the data analysis methods, and tuning the chemometric methods to adapt to these data features. These efforts will transform a time-consuming process that is semi-quantitative at best into a rapid and efficient method that will provide precise quantitative results for biomarker identification. The broader impacts of the work lie in providing biological scientists with the tools to characterize the amounts and relationships of various molecular players as they vary depending on growth, disease state and environmental exposure. This research will contribute to achieving a holistic level of understanding of complex biological systems, which in turn has tremendous promise for understanding and curing disease, enhancing agriculture and preserving the environment.

This is a collaborative project being carried out with Prof. Peter W. Carr at the University of Minnesota and Prof. Dwight R. Stoll at Gustavus Adolphus College.


Virginia Commonwealth University | Chemistry Department
Contact Me
Date Last Modified: April 22, 2013