Automated and rapid bacterial identification using LC-mass spectrometry with a relational database management system

• Published in: Proceedings. 2004 IEEE Computational Systems Bioinformatics Conference, 2004. CSB 2004 pp. 472 - 473
• Main Authors: Deshpande, S.V., Jabbour, R.E., Wick, C., Snyder, A.P.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 2004
• Subjects: Application software; Data analysis; Deconvolution; Mass spectroscopy; Microorganisms; Proteins; Purification; Relational databases; Software algorithms; Software development management
• ISBN: 9780769521947; 0769521940
• DOI: 10.1109/CSB.2004.1332463

We have developed an integrated and automated software application for rapid bacterial identification using a relational database management system and liquid chromatography-electrospray-ion trap mass spectrometry (LC-ESl-MS). LC-ESI-MS is used to generate chromatographic profiles of proteins in a bacterial sample along with a software program that automates the data analysis. The software program ProMAPDB automates the data collection, peak identification, spectral purification, mass spectral integration of scans in a peak, and assignment of molecular weights for observed proteins by using a deconvolution algorithm described by Zhang and Marshall. The approach generates a list of biomarker masses along with retention time and relative abundance for all masses obtained by the algorithm. The list of masses is stored in a relational database as a reference library including the sample information such as growth conditions and experimental information. The identification of unknown samples is performed by correlation to the relational database. The bacterial database includes E. coli, Bacillus subtilis, B. thuringiensis, and B. megaterium. The approach has been tested for bacterial discrimination and identification from the mass spectra of mixtures of microorganisms and from mass spectra of organisms at different growth conditions. Experimental factors such as sample preparation, reproducibility, mass range and mass accuracy tolerance are also addressed and evaluated. This approach has the potential for reliable and accurate automated data analysis.