PanBGC: A pangenomic framework for systematic analysis of biosynthetic gene cluster diversity

Abstract

Microbial secondary metabolites are critical sources of antibiotics, anticancer agents, and otherbioactive compounds, with over 60% of approved drugs derived from natural products. Thesecompounds are encoded by biosynthetic gene clusters (BGCs), which contain the geneticinstructions for producing complex chemical structures. While computational tools can identifyand group BGCs into families based on similarity, current approaches lack systematicframeworks for analysing the internal genetic diversity that drives chemical innovation withinthese families. This limitation represents a significant gap in understanding how biosyntheticpathways evolve and generate the remarkable chemical diversity observed in microbial naturalproducts.In my PhD project, I developed PanBGC, a computational framework that adapts pangenomicprinciples to analyse biosynthetic gene cluster diversity. Just as pangenomics that analysesmicrobial genomic diversity within species populations, PanBGC treats gene cluster familiesas structured populations rather than isolated genomic islands. Applied to over 250,000 BGCsfrom more than 35,000 microbial genomes, representing over 80,000 gene cluster families, thePanBGC framework revealed patterns of genetic organization within biosynthetic families. Themost significant finding was the identification of contrasting evolutionary dynamics: while mostBGC families exhibit closed gene repertoires with limited acquisition of entirely novel genes,they simultaneously demonstrate high compositional plasticity in how existing geneticcomponents of one family are combined within individual clusters. This finding indicates thatbiosynthetic innovation operates primarily through modular reorganization of evolutionarilyvalidated genetic components rather than through continuous incorporation of novel geneticmaterial.Functional analysis revealed that core genes were predominantly associated with essentialenzymatic activities required for basic metabolite biosynthesis, while accessory genes wereassociated with tailoring reactions and regulatory functions that contribute to structuraldiversification. To make these analyses accessible to the research community, PanBGC-DBwas developed as an interactive web platform containing precomputed analyses of genecluster families, enabling systematic investigations of BGC family diversity.The framework addresses practical applications in natural product research. For genomemining, the systematic organization enables prioritization of clusters with unusual genecombinations that may represent novel chemical scaffolds. In synthetic biology, theidentification of accessory genes that co-occur with specific core pathways provides evidencefor functional compatibility, improving the success rate of biosynthetic engineering comparedto traditional trial-and-error approaches.The results of this PhD thesis establish a systematic framework for understandingdiversification within BGC families. This work provides both fundamental insights intobiosynthetic evolution and practical tools for natural product discovery, representing a shifttoward population-level thinking in biosynthetic gene cluster analysis.