Abstract
We present a novel informatics system for automatic high-throughput processing of metabolic profiling data obtained from liquid chromatography-Fourier transform ion cyclotron resonance-mass spectrometry (LC/FT-ICR-MS). The data processing system is implemented with tools that facilitate 1) normalization of m/z value fluctuations in each single scan using a regression analysis of m/z values of internal mass calibrants, 2) organizing corrected m/z values from a large number of scans in a single m/z value-matrix, and 3) elimination of noise peaks and detection of statistically significant peaks. In addition, we report an application of the system to metabolome data obtained from tomato fruit samples. Peaks selected by the system are annotated with chemical information including putative molecular formulas, MS/MS fragmentation patterns, and UV/VIS absorption patterns. Our data processing system provides firm basis for metabolite annotation procedure, and allows coherent analysis of numbers of LC/FT-ICR-MS data.