Comparative BAC ‐based physical mapping of O ryza sativa ssp. indica var. 93–11 and evaluation of the two rice reference sequence assemblies

• Published in: The Plant journal : for cell and molecular biology Vol. 77; no. 5; pp. 795 - 805
• Main Authors: Pan, Yonglong, Deng, Ying, Lin, Haiyan, Kudrna, David A., Wing, Rod A., Li, Lijia, Zhang, Qifa, Luo, Meizhong
• Format: Journal Article
• Language: English
• Published: 01-03-2014
• ISSN: 0960-7412; 1365-313X
• DOI: 10.1111/tpj.12412

Summary
Reference sequences are sequences that are used for public consultation, and therefore must be of high quality. Using the whole‐genome shotgun/next‐generation sequencing approach, many genome sequences of complex higher plants have been generated in recent years, and are generally considered reference sequences. However, none of these sequences has been experimentally evaluated at the whole‐genome sequence assembly level. Rice has a relatively simple plant genome, and the genome sequences for its two sub‐species obtained using different sequencing approaches were published approximately 10 years ago. This provides a unique system for a case study to evaluate the qualities and utilities of published plant genome sequences. We constructed a robust
BAC
physical map embedding a large number of
BAC
end sequences forrice variety 93–11. Through
BAC
end sequence alignments and tri‐assembly comparisons of the 93–11 physical map and the two reference sequences, we found that the Nipponbare reference sequence generated using the clone‐by‐clone approach has a high quality but still contains small artifact inversions and missing sequences. In contrast, the 93–11 reference sequence generated using the whole‐genome shotgun approach contains many large and varied assembly errors, such as inversions, duplications and translocations, as well as missing sequences. The 93–11 physical map provides an invaluable resource for evaluation and improvements toward completion of both
N
ipponbare and 93–11 reference sequences.