Abstract
Distinct breeding populations of Coffea canephora often exhibit genetic divergence, yet the biological pathways underlying yield and leaf rust resistance in contrasting populations remain poorly understood. Here, we performed a comparative genomic analysis of two populations (Premature and Intermediate) to dissect the genetic architecture of coffee bean production, green bean yield, and leaf rust incidence. By integrating single-SNP association, machine learning (Bootstrap Forest), and Gene Ontology (GO) pathway analysis, we found that the Premature population's traits were linked to specialized metabolic pathways, particularly lipid modification and organelle lumen-associated processes. In contrast, the Intermediate population was governed by core cellular machinery, with significant enrichment for actin cytoskeleton regulation and salicylic acid signaling. These findings demonstrate that distinct breeding populations achieve agronomic success through fundamentally different biological strategies and provide a reusable resource of ranked SNP lists for targeted, population-aware breeding.