Abstract:Abstract: Objective To construct the matrix-assisted laser desorption ionization time-of-fight mass spectrometry ( MALDI-TOF MS)in-house database and evaluate its application value in the identifceation of non- tuberculous mycobacterium ( NTM ). Methods Seventy-two NTM clinical isolates were ollected during January 2018 and May 2021, and were identifed by the direct homologous sequencealignment and MALDI-TOF MS, respectively. The MALDI-TOF MS method for the idenifcation of NTM was further divided into the experimental group and control group,which was based on the " in-house database + MBT DB5989 database" and " MBT DB5989 data- base",respectively. The identification coincidence rate of the two groups was compared and the application efficiency of MALDI-TOFMS in-house database for the identification of NTM was evaluated. Results Using the results of molecular identification as the gold standard, the coincidence rates of the control group and experimental group were51 .61% (32/62) and 95. 16% (59/62), respectively, and the ltter was signifcantly higher than the former (x2 = 30.102, P<0.001 ). After removing the NTM stains missing in the MBT DB5989 database, the coincidence rates of the control group and experimental group were 86.48% ( 32/37) and 97.30% (36/37), respectively, with no statistical significance between them (X2 =2.902, P=0.088 ). One strain of Mycobacterium paragordonae in the experimental goup was inorrectly identified as Mycobacterium gordonae. Conclusion The selection of MALDI-TOF MS databases directly affects the identification results of NTM, and the in-house database can improve the efficiency of MBT DB5989 database for the identification of NTM.