Fig. 5

Comparative genomic analysis of circum-basmati rice evolution. The Oryza sativa variety groups are labeled as circum-aus (cA), circum-basmati (cB), indica (I), and japonica (J), and the wild relative is O. rufipogon (R). a Maximum likelihood tree based on fourfold degenerate sites. All nodes had over 95% bootstrap support. b Percentage of genes supporting the topology involving japonica Nipponbare, circum-basmati Basmati 334 (B334) and Dom Sufid (DS), and O. rufipogon after an Approximately Unbiased (AU) test. c Results of ABBA-BABA tests. Shown are median Patterson’s D-statistics with 95% confidence intervals determined from a bootstrapping procedure. For each tested topology, the outgroup was always O. barthii. d Percentage of genes supporting the topology involving circum-aus N22, circum-basmati, and indica R498 after an AU test. e Per-chromosome distribution of D-statistics for the trio involving R498, N22, and each circum-basmati genome. Genome-wide D-statistics with 95% bootstrap confidence intervals are indicated by the dark and dotted lines. f Model of admixture events that occurred within domesticated Asian rice. The direction of admixture has been left ambiguous, as the ABBA-BABA test cannot detect the direction of gene flow