Establishment of a matrix-assisted laser desorption/ionization time-of-flight mass spectrometry database and evaluation of different methods for cluster analysis of marine bacteria.

Introduction. Accurate and rapid identification of marine bacteria is essential for the precise diagnosis and treatment of infectious diseases caused by marine pathogens. The accuracy of microbial identification using matrix-assisted laser desorption/ionization time-of-flight MS (MALDI-TOF MS) primarily depends on the diversity and number of strains included in the reference database.Hypothesis/Gap statement. We hypothesized that expanding the MALDI-TOF MS database with a broader collection of protein spectral profiles from marine bacteria would significantly enhance identification accuracy for these strains.Aim. This study aimed to establish a marine bacterial database to improve identification accuracy and to evaluate the applicability of MALDI-TOF MS-based cluster analysis for tracing the origin of clinical strains.Methodology. We collected 203 strains isolated from marine environments and clinical samples, acquired their MALDI-TOF MS spectra and constructed a mass spectral database specific to marine bacteria. To validate the accuracy of the expanded database, 80 external strains were subsequently tested. Furthermore, we assessed the strain classification efficacy of MALDI-TOF MS cluster analysis and phylogenetic trees constructed from gene sequences.Results. The species-level identification rate increased from 88.75 to 97.5%. The proportion of strains achieving a reliable identification score (>2.3) rose markedly from 43.75 to 91.25%. Cluster analysis based on MALDI-TOF MS demonstrated high accuracy in grouping bacteria at the species level. In addition, the maximum likelihood (ML) phylogenetic tree exhibited significantly higher bootstrap support values compared to the neighbour-joining tree.Conclusion. The expanded marine bacterial database markedly enhances the accuracy and reliability of MALDI-TOF MS for identifying marine pathogens. For species identification and traceability, we recommend a combined strategy that includes initial MALDI-TOF MS screening and verification with phylogenetic trees based on the ML method.