Genus-wide comparative genomics reveals extensive unexplored biosynthetic diversity in Pseudonocardia

Pseudonocardia is an ecologically diverse actinobacterial genus recognized for its ability to produce specialized metabolites. However, its biosynthetic diversity has not been systematically examined at the genus level. In this study, we analyzed 119 quality-filtered Pseudonocardia genomes using comparative genomics, integrating pangenome analysis, average nucleotide identity (ANI), multi-tool biosynthetic gene cluster (BGC) prediction, self-resistance profiling, and gene cluster family analysis. The genus exhibited an open pangenome comprising 37,579 gene families, dominated by cloud genes (81.24%) with a relatively small core genome (3.46%), indicating extensive accessory genome diversity. ANI analysis also identified several inconsistencies between public species assignments and genome-based relationships. Comparative genome mining with antiSMASH, DeepBGC, and GECCO showed widespread biosynthetic potential across all genomes, with DeepBGC identifying 5,150 candidate BGCs at a confidence threshold of ≥ 0.7, substantially expanding the biosynthetic landscape beyond rule-based prediction alone. Ecological comparisons revealed significant differences in BGC abundance and cloud-gene content among habitat groups. After stringent filtering, 180 high-confidence candidate BGCs were identified, including 138 highly uncharacterized regions lacking antiSMASH overlap and 42 candidates belonging to MIBiG-distant gene cluster families. ARTS analysis indicated that self-resistance features are most informative when used alongside complementary genome-mining approaches for candidate prioritization. Overall, our results highlight Pseudonocardia as a rich and still underexplored source of specialized metabolites and provide a comparative genomics workflow for prioritizing genomes and BGCs for future natural product discovery.