Abstract
<jats:p>The genus Trichoderma comprises ecologically and biotechnologically important fungi widely used in agriculture, industrial biotechnology, and biological control. However, publicly available genomes have revealed substantial taxonomic inconsistencies across the genus. Here, we present a comprehensive phylogenomic and comparative genomic analysis integrating one of the largest Trichoderma genome collections analyzed to date. Phylogenomic reconstruction based on 920 conserved single-copy orthologous genes recovered four major evolutionary clades with strong statistical support and revealed widespread taxonomic misannotation affecting multiple species complexes, including T. harzianum, T. asperellum, T. viride, and T. longibrachiatum. Comparative analyses demonstrated marked clade-associated differences in genome size, GC content, repetitive DNA proportion, gene content, and whole-genome conservation patterns. Genome size correlated positively with repetitive element accumulation and gene counts, whereas GC content displayed a negative correlation with assembly size. Additionally, we describe a novel Colombian isolate clustering within a highly supported and divergent lineage with three inconsistently annotated public genomes. This lineage, provisionally designated Trichoderma sp. "CB2", formed a sister group to the Longibrachiatum complex, exhibiting strong internal genomic cohesion and clear divergence from neighboring lineages. Orthology-based analyses identified lineage-specific proteins associated with transcriptional regulation, plant biomass degradation, secondary metabolism, and detoxification pathways. This study provides a phylogenomic framework for the genus Trichoderma and highlights the extent of taxonomic inconsistencies in public genomic databases.</jats:p>