Developing an in silico Rice Genebank

The highest informational content that can be added to a germplasm record will be its complete genome with functional annotation and attribution to traits to bridge the genotype to phenotype divide.  The technologies of high-density SNP discovery, genotyping, and re-sequencing allow this dream to be achievable in the next years. IRRI working with Cornell and other partners in the Rice SNP consortium have undertaken the first two activities for genotyping over 2000 diverse rice types with 1M SNPs. The SNP pool for array design was derived from resequencing over 150 genomes. Another outcome is the production of reference-guided de novo assemblies for those with data at 30x plus depth such as IR 64 (see Mauleon et al). For the resequencing activity, a project by BGI, CAAS and IRRI will sequence 3,000 genomes in phase one.  The 2K SNP and 3K sequencing sets are in line for phenotyping under the global phenotyping network.  All activities are embedded in the Global Rice Science Partnership and will be in the public domain.  The impending data deluge requires rethinking data management with tracking at the sample level for descendants of the conserved accession. Approaches to integrate wide scale genomic data with genebank data will involve federated approaches. Tools built on this interface to query regions filtered by local structure will target SNP/marker discovery for breeding.