The common bean (Phaseolus vulgaris L.) is an important grain legume crop [1,2] whose life history offers an ideal evolutionary model to identify adaptive variants suitable for breeding programs [3]. Here we present the first common bean pan-genome based on five high-quality genomes and whole-genome reads representing 339 genotypes. We found ~243 Mb of additional sequences containing 7,495 protein-coding genes missing from the reference, constituting 51% of the total presence/absence variations (PAVs). There were more putatively deleterious mutations in PAVs than core genes, probably reflecting the lower effective population size of PAVs as well as fitness advantages due to the purging effect of gene loss. Our results suggest pan-genome shrinkage occurred during wild range expansion from Mexico to South America, with more PAV loss per individual in Andean vs Mesoamerican populations. Selection during wild spreading and domestication was also associated with PAV loss involved in important adaptive traits. Our findings provide evidence that partial or complete gene loss was a key adaptive trait leading to localized and genome-wide reductions. This departs from established paradigms and reveals how evolutionary forces shape gene composition within plant genomes. The common bean pan-genome offers a valuable resource for legume research and breeding, climate change mitigation, and sustainable agriculture.

Adaptive gene loss in the common bean pan-genome during range expansion and domestication