Toward Smart Databases for Oat Genomics

There exist many database applications and software programs that are useful for managing and analysing crop genomics data. Yet we often find ourselves intimidated by the need to answer seemingly-simple questions from complex and diverse data.  As an example, we wished to interrogate several marker data sets to extract genotypes of oat varieties that contained high similarity to specific parental haplotype blocks that were hypothesized to contain favourable QTL alleles.  Furthermore, we wanted an automatic query that was smart enough to (1) expand the haplotype region to contain additional linked or associated markers and (2) recode the result such that alleles common to a target haplotype were represented in identical phase.  For example, "1111111101" would code the genotype of a variety that was identical to the target haplotype at all but one of ten markers.   We have developed a prototype application that can perform this query, and we propose to develop this application further and encourage its integration with other data management platforms.   We will illustrate the utility of this query by extracting putative carriers of favourable "historical" QTL alleles from a large database of SNP and GBS markers developed recently by the Collaborative Oat Research Enterprise (CORE).  The query can be used in both mapping data and diversity data.  As expected, in diversity data it will often identify haplotype blocks in varieties with common ancestry.  We have validated this by comparing query results to ancestral information from the Pedigrees of Oat Lines (POOL) database.  However, this query may also identify near-identical haplotypes that have been maintained through selection in varieties where common ancestry was not documented.  Inspection of these query results may assist in rationalization of historical QTL discovered by bi-parental mapping with those found by association mapping.  Furthermore, we propose that `smart queries` may assist in molecular breeding by facilitating the deliberate enhancement of a genomic selection strategy through identification of known QTL haplotypes.