Comparative Analysis of Codon Usage Bias in Nuclear Genome Coding Sequences (CDSs) Across 30 Angiosperm Species Centered on the Leguminosae (Fabaceae)

Synonymous codon usage bias (CUB) modulates gene expression by influencing translational efficiency and accuracy, shaped by mutational biases and natural selection. A unified framework is essential to identify lineage-specific codon preferences and optimize expression and sequence design across species.

Nuclear CDSs from 30 angiosperm species, with a focus on Fabaceae, were analyzed using CodonW to compute gene-level indices (GC/GC3s, ENc, CAI, CBI, Fop). Codon preference patterns were examined through hierarchical clustering, PCA, and analyses of mutational and selective influences using ENc-GC3s plots, neutrality plots, and PR2-bias.

Substantial variation in GC3s, ENc, and CAI/CBI/Fop was observed, with patterns aligned with phylogenetic differentiation. PCA revealed clear lineage-level clustering based on codon preferences, with Fabaceae species distinctly separated from Poaceae and other groups. Strong correlations between GC and GC3s, as well as between GC and CBI/Fop, highlighted the pivotal role of base composition in shaping codon preference. Additional analyses indicated that selective pressures, alongside mutational biases, contribute to codon usage, with expression-related indices providing evidence of translation-associated selection. This study establishes a comprehensive framework for understanding codon usage across 30 angiosperms, offering valuable insights for codon optimization and heterologous expression in the molecular breeding of legume crops and other species.