Genetic diversity and population structure of Coffea arabica in Saudi Arabia revealed by whole-genome resequencing.

Gene banks are essential for conserving plant genetic resources, yet limited genomic characterization constrains their effective use in breeding. Coffea arabica L., an economically important crop in Saudi Arabia, remains poorly characterized at the genome-wide level. We performed whole-genome resequencing of 78 Saudi C. arabica accessions conserved at the Ministry of Environment, Water and Agriculture (MEWA) Seed Center and PGR Bank, generating 86 sequencing libraries, including eight technical replicates. Reads were aligned to the Cara_r6 reference genome, yielding 468, 133 variants (412, 043 SNPs and 56, 090 indels), from which 17, 452 linkage disequilibrium-pruned SNPs were retained for downstream analyses. We identified a pervasive cross-mapping artefact between the homoeologous subgenomes of this allotetraploid species, affecting 63.7% of markers and artificially inflating observed heterozygosity. Correcting this artefact reversed the inbreeding coefficient from negative (FIS ≈ -0.60) to positive (FIS = 0.37-0.52), consistent with the predominantly self-pollinating biology of C. arabica. Population structure analysis identified seven ancestry groups (K = 7) that were highly robust to artefact correction (Adjusted Rand Index = 0.93) but showed no correspondence with collection regions. Analysis of molecular variance attributed only 0.94% of the total genetic variation to differences among regions (p = 0.206). We developed a candidate 150-SNP fingerprinting panel with high marker informativeness (mean PIC = 0.374); however, independent validation is required before routine application because closely related accessions could not always be distinguished. These findings provide a robust genomic framework for the conservation and management of Saudi coffee genetic resources and demonstrate the importance of subgenome-aware analytical approaches for whole-genome studies of allotetraploid C. arabica.