Abstract View
An Automated Approach to Identify and Quantify Compounds in Large GC-MS Datasets using Positive Matrix Factorization
SOHYEON JEON, Michael Walker, Claire Fortenberry, Brent Williams, Washington University in St. Louis
Abstract Number: 478
Working Group: Instrumentation and Methods
Abstract
Gas chromatography-mass spectrometry (GC-MS) has long been applied to identify and quantify individual chemical compounds in environmental samples. Typically, users identify each chemical compound by comparing mass spectral fragmentation patterns with known standards or MS libraries and quantify by integrating peaks to compare the variability between samples. Software that can automatically carry out these processes has been developed, but often requires manual quality control inspection of peaks. This can be particularly time-consuming for peaks with large retention time shifts or low abundance. In addition, this process may generate errors by the researcher’s subjective correction. We evaluate here an automated approach of identifying and quantifying analytes of interest in the chromatogram for large dataset by using Positive Matrix Factorization (PMF) based on peak information obtained from Igor-based automated single-ion peak fitting method. To implement our automated analysis, a small sub-section of each chromatogram, corresponding to the retention time of an analyte of interest, is determined automatically based on a reference chromatogram before running PMF. Then, the PMF analysis generates solutions with increasing number of factors until a set of criteria related to the compound’s mass spectrum and peak shape are met. With this approach, users can automatically implement PMF for all analytes of interest with minimal effort. Previously quantified peaks (compounds) from the Air Composition and Reactivity from Oudtoor aNd Indoor Mixing (ACRONIM) campaign are utilized to assess the strengths and limitations of this approach. This technique shows potential to improve data processing efficiency to identify and quantify analytes of interest in chromatograms from a range of environmental samples.