Identification of seeds and mislabeled samples from seed banks
Considering that the rice-specific barcode resolved almost all rice species, we used it to identify 53 accessions of seeds from seed banks or field collections. Nine (17%) mislabeled samples were found (Fig. S6). These samples were all from species-rich genome types Aand C . This was not surprising, as in the A genome type, there is still some confusion between O. rufipogon andO. nivara , and between O. glaberrima (O. barthii ) and O. glumipatula . Similarly, in the C genome type, there is confusion between diploid and tetraploid O. punctata , and among tetraploid O. alta , O. latifolia , and O. minuta .