MFPD: A Multiple Fungal Pathogen Detection Pipeline Across Diverse Habitats

Fungal pathogens threaten the health of humans, animals, and plants. ITS sequencing offers an effective approach for detecting fungal pathogens; however, a comprehensive pathogen database and associated tailored pipeline are still lacking. This study introduces the multiple fungal pathogen detection (MFPD) pipeline, which incorporates an accurate and high-speed sequence alignment algorithm for broad-habitat pathogen identification. The curated MFPD database includes 95 660 full-length ITS sequences from 4924 reported fungal pathogen species. In silico experiments show that the full-length ITS achieves the highest accuracy in pathogen detection (average 99.34%), outperforming both the ITS1 and ITS2 subregions. Benchmarking against existing tools, including FUNGuild, FungalTraits, and ISHAM-ITS, shows that MFPD achieves the highest F1 scores in mock communities (0.89 for both plant and human-animal pathogens) and detects the broadest spectrum of pathogenic taxa in real samples. In addition to identifying causal pathogens, MFPD can also detect coinfecting pathogens in biological and environmental samples. Together, our work supports pathogen surveillance across diverse sectors, including clinical, agricultural, and livestock systems within a One Health framework.