Abstract
In order to elucidate the genome evolution of the Asian cultivated rice, Oryza sativa, we compared the complete genome sequence with BAC end sequences (BESs) of two wild rice species, O. rufipogon and O. nivara. We also mapped the BESs of the African cultivated rice species, O. glaberrima, which was used as the outgroup species. A phylogenetic tree constructed on the basis of the multiple alignment of non-coding sequences suggested that O. sativa was derived from common ancestor of O. rufipogon and O. nivara. By calculating the extent of the nucleotide divergence between O. sativa and other species over the entire genome, we found that some regions showed a significantly lower divergence level than expected. Further analysis based on the maximum parsimony method could reveal that at least 12 regions making up 4.8% of the O. sativa genome showed a tree topology incongruent with the species tree. This may be due to long-standing ancestral polymorphisms or gene introgression events that occurred about 0.1 MYA between the cultivated and wild rice species.