November 13 - 16, Charleston, SC, USA, SERMACS 2002

Enabling Benchtop Chemists to Analyze MS Data: Delivery of Complex Chemometric Algorithms, Auto-fragmentation and Assignment to the Desktop

Antony J. Williams, Vitaly Lashin and Mark Bayliss

Abstract

Data creation today is highly automated. The result is an abundance of data available to a chemist in aid in the characterization of their samples. The various forms of MS data, including in its hyphenated form delivers a complex array of data for interpretation. The chosen path of late has commonly been to deliver molecular weight information for a compound rather than a mass spectrum with fragmentation detail. Fragmentation carries more information regarding structural identification and should be used where possible but this requires educating chemists in the detailed interpretation of MS spectra, a task commonly left to the professional spectrometrist.

Similarly, LC/MS is commonly available today for chemists to access. Commonly there can be issues with extracting the information from the LC/MS data due to low signal-to-noise ratios. Scientists at Eastman-Kodak previously reported the development of a component detection algorithm, CODA, for noise and background reduction in LC/MS chromatograms. Recently a second algorithm, COMPARELCMS, was developed to distinguish dissimilarities in CODA reduced chromatograms permitting the extraction of component differences in similar LC/MS data sets.

We have developed a single software solution which provides access to tools for the facile interrogation and both manual and automated assignment of mass spectral data. We have also delivered easy-to-use chemometrics capabilities to aid in the analysis of LC-MS data. These processing tools support automated data reduction for open access LC/MS and batched LC/MS samples from parallel synthesis and are fully integrated with data capture, structure assignment, reporting, and databasing.