Optimizing filamentous fungi identification by MALDI-TOF MS: a comparative analysis of key factors

Abstract

Purpose: The identification of filamentous fungi in clinical microbiology laboratories remains a challenging task. Although matrix-assisted laser desorption/ionization–time of flight mass spectrometry (MALDI-TOF MS) has revolutionized microbial diagnostics by enabling rapid and accurate species-level identification, its application to molds is still evolving. This study aims to evaluate the performance of two Bruker MALDI-TOF MS systems, Sirius One and Microflex 3.1, for the identification of filamentous fungi using different extraction protocols and database configurations. Method: A total of 68 filamentous fungal isolates, including clinically significant species, were analyzed. Fungal cultures were processed under standardized conditions using two protein extraction methods: a detailed in-tube extraction with ethanol, formic acid, and acetonitrile, and a direct on-plate extraction. Spectra were acquired using both Sirius One and Microflex 3.1 systems, and identifications were performed using manufacturer-provided databases and the MSI-2.0 database. Results: The Sirius One system outperformed Microflex 3.1, achieving a 92.6% correct identification rate with the MSI-2 database compared to 70.6% for Microflex (p < 0.01). When using manufacturer-provided databases, identification rates were lower: 51.5% for Sirius One and 41.2% for Microflex. Notably, the on-plate extraction method performed comparably to the in-tube method, achieving 94.1% accuracy with Sirius One and the MSI-2 database. Conclusion: The combination of the Sirius One system, MSI-2.0 database, and on-plate extraction method provides a highly effective and time-efficient workflow for the identification of filamentous fungi in routine clinical diagnostics, reaching 94.1% accuracy. This approach is recommended for implementation in clinical mycology laboratories, though further optimization of manufacturer-supplied databases remains necessary. © The Author(s) 2025.