%0 Journal Article
%T Model-based peak alignment of metabolomic profiling from comprehensive two-dimensional gas chromatography mass spectrometry
%A Jaesik Jeong
%A Xue Shi
%A Xiang Zhang
%A Seongho Kim
%A Changyu Shen
%J BMC Bioinformatics
%D 2012
%I BioMed Central
%R 10.1186/1471-2105-13-27
%X Using experimental data of a mixture of metabolite standards, we demonstrated that our method has better performance than other existing method which is not model-based. We then applied our method to the data generated from the plasma of a rat, which also demonstrates good performance of our model.We developed a model-based peak alignment method to process both homogeneous and heterogeneous experimental data. The unique feature of our method is the only model-based peak alignment method coupled with metabolite identification in an unified framework. Through the comparison with other existing method, we demonstrated that our method has better performance. Data are available at http:/ / stage.louisville.edu/ faculty/ x0zhan17/ software/ software-development/ mspa webcite. The R source codes are available at http://www.biostat.iupui.edu/~ChangyuShen/CodesPeakAlignment.zip webcite.2136949528613691Comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry (GCxGC/TOF-MS) has been employed in analysis of complex samples in metabolomics studies, especially in cancer [1-3]. However, experiment output still suffers from substantial variations, challenging the data interpretation in these studies. For this reason, the methodological development at the data analysis stage has become a crucial research area, which is still at its infancy. Throughout the paper, we mean experiment by technical run, which is performed by machine.In practice, there are two types of experiment variations: variation within an experiment and variation between experiments. The need to control variation within an experiment has been reported by many researchers [4-6]. Theoretically, all instrument signals generated by one metabolite species should be reported as a single peak within an experiment. However, in reality, multiple peak entries can occur due to the abnormality in peak shape, high sensitivity of the peak detection algorithm, and experimental cause such as short modulatio
%U http://www.biomedcentral.com/1471-2105/13/27