Fig. 1

Origin and population structure of African and Asian rice Oryza sativa: a Geographic origin and sampled number of accessions is shown from both continents. b PCA of 2710 rice accessions from Africa and Asia, along axes 1 and 2. The first axis (39% of the variance) differentiates japonica (GJ-trp, GJ-sbtrp, GJ-tmp, and GJ-adm) from indica (XI-1A, XI-1B, XI-2, XI-3, and XI-adm). The cAus ecotype differs from other groups on the second axis, which explains 21% of the total variance. Basmati accessions (cBas), very close to Japonica, appear to belong to the GJ group but received a contribution from cAus ecotype. c Unsupervised hierarchical clustering based on 3023 accessions. Each bar represents an accession. The colours indicate a genetic cluster obtained for K values ranging from 2 to 10. K values from 2 to 9 (but not k = 8, because no significant cluster were observed) are shown. K = 9 has the lowest cross-validation error value. Clusters were assigned to varietal groups as described in Wang et al. 2018a, b. The colours blue, red and yellow correspond to the 3 reference clusters japonica, indica and circum-aus respectively. d Cross-validation (CV) error in ADMIXTURE model-based clustering. From K = 2 to K = 10, K = 9 have the lowest error