Title :
Charge Prediction of Lipid Fragments in Mass Spectrometry
Author :
Schrom, B. ; Kangas, L. ; Ginovska, B. ; Metz, T. ; Miller, J.
Author_Institution :
Pacific Northwest Nat. Labratory, Richland, WA, USA
Abstract :
An artificial neural network is developed for predicting which fragment is charged and which fragment is neutral for lipid fragment pairs produced from a liquid chromatography tandem mass spectrometry simulation process. This charge predictor is integrated into software developed at PNNL for in silico spectra generation and identification of metabolites known as Met ISIS. To test the effect of including charge prediction in Met ISIS, 46 lipids are used which show a reduction in false positive identifications when the charge predictor is utilized.
Keywords :
biology computing; chromatography; mass spectra; molecular biophysics; neural nets; software engineering; Met ISIS; PNNL; artificial neural network; charge prediction; charge predictor; false positive identifications; in silico spectra generation; lipid fragment pairs; lipid fragments; lipids; liquid chromatography tandem mass spectrometry simulation process; metabolites identification; Artificial neural networks; Atomic measurements; Databases; Lipidomics; Testing; Training; Vectors; artificial neural network; lipid; machine learning; mass spectrometry;
Conference_Titel :
Machine Learning and Applications and Workshops (ICMLA), 2011 10th International Conference on
Conference_Location :
Honolulu, HI
Print_ISBN :
978-1-4577-2134-2
DOI :
10.1109/ICMLA.2011.45
