Oat Molecular Toolbox: Toward Better Oats - PowerPoint Presentation

1. Oat Molecular Toolbox:Toward Better OatsNick Tinker, 2014-March-5Agriculture and Agri-Food Canada

2. Collaborative Oat Research Enterprise* Mexico:Julio HuertaEduardo Villa senior MirEduardo Espitia

3. What makes oats different, which differences make a better oat ?3

4. CORE Concept Summary4Representative DNA SequenceDiscover genetic differences among oat varietiesDiverse germplasm Marker assays + gene databaseMapping germplasmBreeder GermplasmEvaluate field & seed TraitsGenotype / Trait databaseConsensus mapAnalyse population structure Associate markers with traitsBreeding assays + genomic selection

5.

6. CORE Concept Summary6Representative DNA SequenceDiscover genetic differences among oat varietiesDiverse germplasm Marker assays + gene databaseMapping germplasmBreeder GermplasmEvaluate field & seed TraitsGenotype / Trait databaseConsensus mapAnalyse population structure Associate markers with traitsBreeding assays + genomic selection

7. This is what we were looking forATFunctional differenceSingle Nucleotide Polymorphism = SNPSNP = Marker= Gene LocusAllele

8. SNP assays:8 GTACCATGATCGCTAACTGGCATGGCTTACGGCTTGAC(A) ..................G...................(B) ..................G...................(C) ..................A...................(D) ..................A...................(E) ..................G...................A SNP is a SNP …. no matter how you find it !“Old” non-sequence-based methods (AFLP, DArT)Discover by sequence / assay by design (Illumina Array) - 6000Discover and assay by sequencing (GBS)

9. 6K SNP array annotations9Estimated chromosome (relative to Oliver et al. 2013)Estimated map position (cM relative to Oliver et al. 2013)Flag=1 for framework marker (relative to Oliver et al. 2013)Best match to Brachypodium distachyon genome Best match to Oryza sativa genome Best match to Hordeum vulgare genomeBest BLAST description from BLAST2GO Minimum E value from BLAST2GO Gene Ontology terms from BLAST2GO Enzyme Code from BLAST2GO Protein Accession (NCBI) from SNPmetaShort protein name of best match from SNPmetaPredicted SNP position in coding sequence from SNPmetaPredicted SNP position in codon from SNPmetaPredicted codon for allele 1 from SNPmetaPredicted codon for allele 2 from SNPmetaPrediction if amino acid change is silent, from SNPmetaPredicted amino acid for SNP allele 1 from SNPmetaPredicted amino acid for SNP allele 2 from SNPmeta38 important annotations (consolidated from many more)SNP Locus NameSNP Discovery method (from Table 1)Reason for inclusion / in-silico predicted performance aBead Type (1 a transition SNP, 2=transvesion SNP)Illumina design scoreSuccessful conversion to 6K BEAD assaySuccessful assay based on MAF>0 and H<=10 in 595 progenySNP bases (A,T,G,C in format [A/T] ) SNP design sequence (SNP in square brackets)Full contig sequence (for SNPs called from an assembly)Comments made in GenomeStudio Genotyping ModuleComments made in GenomeStudio Clustering ModuleNumber of clusters formed in Clustering ModuleIllumina Gentrain scoreNon-missing calls across 1055 progeny (%)Number of AA (alleles 1) calls in 595 breeding linesNumber of BB (alleles 2) calls in 595 breeding linesPercentage of AB calls in 595 breeding linesMinor Allele Frequency in 595 breeding lines

10. Developing functional gene assays…..10Eric Jackson et al. (unpublished)

11. CORE Concept Summary11Representative DNA SequenceDiscover genetic differences among oat varietiesDiverse germplasm Marker assays + gene databaseMapping germplasmBreeder GermplasmEvaluate field & seed TraitsGenotype / Trait databaseConsensus mapAnalyse population structure Associate markers with traitsBreeding assays + genomic selection

12. 12

13. CORE Concept Summary13Representative DNA SequenceDiscover genetic differences among oat varietiesDiverse germplasm Marker assays + gene databaseMapping germplasmBreeder GermplasmEvaluate field & seed TraitsGenotype / Trait databaseConsensus mapAnalyse population structure Associate markers with traitsBreeding assays + genomic selection

14. The consensus map challengeConsensus map is an abstraction Smooth out errors in component maps Put all markers on one mapFind ‘most popular order’ when real differences exist Why ?Merge information from diverse studiesPlan experimentsOrganize databasePredict optimum genotypes Sequence genome, clone genes, perfect predictions14

15. Building block populations (“component maps”)15PopulationAbbr.Pop. SizeMarker TypeContributed byReferenceGS-7 x BoyerGB766KBonman et al. Babiker et al. in pressProvena x GS-7PGS986K, GBSBonman et al. Babiker et al. in pressProvena x BoyerPB1396KBonman et al. Babiker et al. in press86-1156 x Clintland 64IL41126KKolb et al. Foresman et al., in press86-6404 x Clintlant 64IL51716KKolb et al.Foresman et al., in pressAssiniboia x MN841801AM1616KMitchell-Fetch et al.Nanjappa et al. in pressOtana x PI269616OP986K, GBSCarson et al.Oliver et al., 2013CDC SolFi x HiFiSH536K, GBSBeattie et alOliver et al., 2013Dal x ExeterDE1456K,GBSTinker et al.Hizbai et al., 2012Hurdal x Z-597HZ536K,GBSBjørnstad et al. Oliver et al., 2013Ogle x TAMO 301OT536K, GBSJackson et al.Portyanko et al., 1995Kanota x OgleKO526K, GBSTinker et al. O'Donoughue et al., 1995

16. High Density Hexaploid Oat Map161C 2C 3C 4C 5C 6C 7C8A 9D 10D 11A 12D 13A14D 15A 16A 17A 18D 19A 20D 21D

17. CORE Concept Summary17Representative DNA SequenceDiscover genetic differences among oat varietiesDiverse germplasm Marker assays + gene databaseMapping germplasmBreeder GermplasmEvaluate field & seed TraitsGenotype / Trait databaseConsensus mapAnalyse population structure Associate markers with traitsBreeding assays + genomic selection

18. Spring and Winter are definitely different:18

19. Model-based analysis reveals structure of 17 different breeding programs / regions19NDSUWinnOttawaNordTexasIdahoModel: K=10 (colours show % of diagnostic alleles)

20. Why does structure matter ?20“Winter” allelesTexas varietiesNorthern Prairie VarietiesSNP and GBS markers“Spring” alleles

21. CORE Concept Summary21Representative DNA SequenceDiscover genetic differences among oat varietiesDiverse germplasm Marker assays + gene databaseMapping germplasmBreeder GermplasmEvaluate field & seed TraitsGenotype / Trait databaseConsensus mapAnalyse population structure Associate markers with traitsBreeding assays + genomic selection

22. Genome Wide Association Mapping (GWAS) Concept is simple: which markers are correlated with a trait which varieties have the good alleles at those loci Hundreds of good predictions from CORE Specialists are refining these predictionsCorrelate with known disease resistanceGenotype x Environment interactionExplore candidate genes22

23. Genomic selectionGive every marker a weightAdvantagesSimple: one abstraction, one inference: “best breeding value”Less likely to be influenced by structure ? DrawbacksTends to improve within a good populationNot good at introducing new allelesArtifacts can go un-noticed 23

24. Integrating multiple inferences24

25. ConclusionsCORE data is a rich foundationAlready supporting new oat science Moving toward a “universal” public oat databaseNow mobilizing to support molecular breedingChallenges:Develop “comfort level” with big-data and abstractionsBuild smart-tools into database (“automated abstractions”)Commit to continue sharing (experience, data and germplasm)Predict crosses, not just selectionsUse tools to access wild relatives25