Whole Genome Profiling provides a robust framework for physical mapping and sequencing in the highly complex and repetitive wheat genome
1 INRA-UBP, UMR1095, Genetics Diversity and Ecophysiology of Cereals, 234 Avenue du Brezet, 63100 Clermont- Ferrand, France
2 Keygene N.V., Agro Business Park 90, 6708 PW, Wageningen, The Netherlands
3 Amplicon Express Inc., 2345 NE Hopkins Ct., Pullman, WA, USA
4 CEA-CNS, 2 rue Gaston Crémieux, CP5706, 91057 Evry cedex, France
5 Department of Plant Pathology, Kansas State University, Manhattan, KS, USA
BMC Genomics 2012, 13:47 doi:10.1186/1471-2164-13-47Published: 30 January 2012
Sequencing projects using a clone-by-clone approach require the availability of a robust physical map. The SNaPshot technology, based on pair-wise comparisons of restriction fragments sizes, has been used recently to build the first physical map of a wheat chromosome and to complete the maize physical map. However, restriction fragments sizes shared randomly between two non-overlapping BACs often lead to chimerical contigs and mis-assembled BACs in such large and repetitive genomes. Whole Genome Profiling (WGP™) was developed recently as a new sequence-based physical mapping technology and has the potential to limit this problem.
A subset of the wheat 3B chromosome BAC library covering 230 Mb was used to establish a WGP physical map and to compare it to a map obtained with the SNaPshot technology. We first adapted the WGP-based assembly methodology to cope with the complexity of the wheat genome. Then, the results showed that the WGP map covers the same length than the SNaPshot map but with 30% less contigs and, more importantly with 3.5 times less mis-assembled BACs. Finally, we evaluated the benefit of integrating WGP tags in different sequence assemblies obtained after Roche/454 sequencing of BAC pools. We showed that while WGP tag integration improves assemblies performed with unpaired reads and with paired-end reads at low coverage, it does not significantly improve sequence assemblies performed at high coverage (25x) with paired-end reads.
Our results demonstrate that, with a suitable assembly methodology, WGP builds more robust physical maps than the SNaPshot technology in wheat and that WGP can be adapted to any genome. Moreover, WGP tag integration in sequence assemblies improves low quality assembly. However, to achieve a high quality draft sequence assembly, a sequencing depth of 25x paired-end reads is required, at which point WGP tag integration does not provide additional scaffolding value. Finally, we suggest that WGP tags can support the efficient sequencing of BAC pools by enabling reliable assignment of sequence scaffolds to their BAC of origin, a feature that is of great interest when using BAC pooling strategies to reduce the cost of sequencing large genomes.