- 1. Theme 2.- Development of improved RTBvarieties: Breeding
ApplicationSelection of superior potato progenitors forrealizing
heterosis supported by high throughputgenotyping and Genomic
selectionMerideth BonierbaleRTB Annual MeetingSept. 29, 2014
2. Heterosis Superior performance of hybrids overparents
Expressed as an increase in biomass,yield, fertility, resistance to
pathogens, ortolerance to climatic stress Maximizing heterosis
implies an increasein heterozygocity and a number of
multi-allelicloci 3. ObjectiveIdentify parents withinpotentially
heterotic genepoolsto exploit heterosis for thedevelopment of high
perfomancetetraploid hybrids 4. Expected outputNew high-yielding
hybrids thatcombine factors for adaptation toresource-limiting
environmentswith end-user preference traitsin reduced time frame 5.
Linkage to RTB FlagshipContributes to accelerate CIP Delivery
FlagshipAGILE, RESILIENT AND PRECOCIOUS POTATOVARIETIES (70-90
days) through the applicationof HTP genotyping, GWAS and S to
expeditegains for yield, earliness, drought and heattolerance, and
long critical photoperiod.Discovery Flagship: Next
GenerationBreeding Systems 6. Research OutcomeNARS and ARI will
have access tosuperior progenitors, abdto tools(MAS, selection
based on GEBV) toselect robust and stable candidatevarieties 7.
Expectation Shprtem breeding cycles in such away that the rate of
genetic gainper unit of time and cost can bezccelerated. 8.
Materials: Two advanced tetraploidpotato breeding populations bred
atCIP Pilot study of HeterosisPopulationB3PopulationLTVR Yield
gainsthroughHeterosis CombiningLate blight +Virus
ResistanceAdaptation tolowland tropicsHeat toleranceResistance
tovirusesNew HybridPopulationLTVRB3Field resistanceto late
blightAdaptation tohighland tropics 9. MethodologyGenotype
population samples using theSolCAP Infinium SNP array (8303
SNP)Apply GWAS to identify markers associatedwith adaptation
traitsApply GS to select best parents based onGEBVGenerate a
prediction model fromtraining populationCompute GEBVsPerform cross
validation of thepredictiona 10. ****B3 Populationn=103LTVR
Populationn=101*Genepool selection: 101 LTVR and 103B3 breeding
lines Genotyped with 52 SSR Neighbor Joining Treeconstructed using
DARWIN LTVR and B3 grouped into two distinct clusters High
diversity is observed within populations 11. TraitsTrait Component
traitsVariableLong criticalphotoperiodTuberinitiationTuber
induction (Scale 1 (the least) -9(the most)Stolon lengthand
number1= very short -9 very long1= few 9 manyBulking (Marketable
tuber number / total tubernumber) * (Number of tuberized
plants/total number of plants) *100HeatToleranceHeat defects 1=no
defects- 9= >80% of tuber defects 12. TraitsTrait
ComponenttraitsVariableDroughttoleranceRoot traits Root
lengthNumber of rootsRoot fresh and dry weightRoot angleStolon
traits Stolon numberStolon diameterStolon fresh and dry weightTuber
traits Tuber numberTuber fresh and dry weightHarvest Index Tuber
Fresh Weight (FW) / Totalbiomass (FW)* X 100*Total biomass= FW(leaf
+ stem+ stolon + tuber +root 13. TraitsTrait
ComponenttraitsVariableDroughttolerancePhysiologicaltraitsChlorophyll
content(SPAD)Canopy temperatureCanopy reflectance(NDVI)Biochemical
Metabolite profiling and NIRS 14. Progress: Genotyping 276 breeding
lines genotyped withSNP 150 LTVR , 50 B3 , and 76
unter-populationhybrids Genotypes assigned to breeding linesusing
fitTetra package of R V1.0(AAAA, AAAB, AABB, ABBB, BBBB) 4738 SNP
markers retained afterquality controlCluster 2 :B3 =37/43LTVR =
5/135LTVRxB3=4/24 15. Population Structure Structure estimated
using 120 SNP (12 /chromosome) in135 LTVR, 43 B3 and 24
inter-population hybrids LTVR and B3 breeding lines were
consistently assignedto Cluster 1 or 2 (some intermixing) while
most LTVR x B3hybrids appeared in the intermixing zone
.1.000.900.800.700.600.500.400.300.200.100.00Cluster 1 :LTVR
=103/135B3 = 4/43LTVRxB3 hybrids=4/24Cluster 2 :B3 =37/43LTVR =
5/135LTVRxB3=4/24Intermixing zoneB3 =4/43LTVR = 27/135LTVRxB3=16/24
16. Phenotyping (Training population =171 breeding
lines454035302520151050Tuber Induction(40 DAPEm)3.0 3.8 4.6 5.4 6.2
7.0 7.8 8.6 9.0Number of breeding linesTuber Induction (1-9)1=
noinduction2= noinduction4= veryweak5= weak 9=
strong45352515454035302520151050Stolon Number(75 DAP)Stolon
Length(75 DAP)2.0 2.9 3.9 4.8 5.8 6.7 7.7 8.6 9.0Number of breeding
linesStolon Length (1-9)A1 5 9B1 5 95-51.8 2.7 3.5 4.3 5.2 6.0 6.8
7.6 9.0Number of breedinglinesStolon Number(1-9) 17. Integrative
Tools for Drought PhenotypingTemperaturedifferencesand NDVI of56
potatoclones Canopy temperature and NDVI may help recognizedrought
tolerant genotypes Leaf temperatures under drought strongly
increasedafter 10 and 20 days after witholding water (DAWW) Drought
strongly affects NDVI after 30 days ofwithholding water 18. GWAS
Model Tested:Mixed Linear Regression y = X + Zg + Considers
structure as co-variate Incorporates kinship to estimate
geneticvariance Statistical package in R 19. GWAS: Manhattan Plot :
AssociatedSNP/chromosome543210-log(p)Stolon length (75 DAP) Long
Days-WarmconditionsChromosomeChr01Chr02Chr03Chr04Chr05Chr06Chr07Chr08Chr09Chr10Chr11Chr1211
significant QTL 20. 543210-log(p)Marketable yield (75 DAP) Long
Days-WarmconditionsChromosomeChr01Chr02Chr03Chr04Chr05Chr06Chr07Chr08Chr09Chr10Chr11Chr12GWAS:
Associated SNP/chromosome2 significant QTL 21. GWAS: Position of
SNP associated with photoperiodresponse and adaptation to warm
conditionsChromosomeTuberInductionStolonNumberStolonlengthBulking75
DAPMarketableyield 75DAPHeat-defectIIIIIIIVVVIVIIVIIIIXXXIXII 22.
GS Model Tested:G BLUP ModelRestricted Maximum Likelihood (REML)
methodsPackage rrBLUP V4.3 of RVariance componentsestimated with 2
dofferentrelationship matrices Additive relationship matrx A model
Euclidian distance Gaussian kernel model 23. Correlation between
Genomic EstimatingBreeding Values (GEBV) and Phenotypic Valuesr
(Additive _ GEBV = 0.89r (Gaussian kernel _GEBV) =
0.941.51.00.50.0-0.5-1.0r (Additive_GEBV) = 0.33r (Gaussian
kernel_GEBV= 0.35Additive based GEBVGaussian kernel basedGEBV0 2 4
6 8 10GEBVPhenotypic
data3.53.02.52.01.51.00.50.0-0.5-1.0-1.5-2.0Additive based
GEBVGaussian kernel basedGEBV0 2 4 6 8 10GEBVPhenotypic
dataInference Populationn=41Training Population n = 130Trait: Tuber
Induction 24. Correlation between GEBV and PhenotypicValuesTraining
Population n=130 Trait : Stolon lengthInference
Populationn=415.04.03.02.01.00.0-1.0-2.0-3.0r (Additive_GEBV) =
0.89r (Gaussian kernel_GEBV) = 0.94Additive based GEBVGaussian
kernel basedGEBV2.52.01.51.00.50 2 4 6 8 10GEBVPhenotypic
data0.0-0.5-1.0-1.5r (Additive _GEBV) = 0.41r (Gaussian
kernel_GEBV=Additive based GEBVGaussian kernel basedGEBV0 2 4 6
8GEBVPhenotypic data 25. Lessons learnedGWAS is a good approach to
study thearchitecture of complex traits and detectunderlying major
genes but is unable tocapture minor gene effectsAllows tagging
genes for thedevelopment of markers to assistselection, as a
complementary tool forbreeding programs 26. Lessons learnedGenomic
Selection addresses smalleffect genes, but several factors
influenceits performancePopulation size affects
predictionaccuracyThe GS model used may not be themost suitable for
predicting GEBV forthe traits under study and other modelsshould be
tested,e.g., Bayesian LASSO 27. Lessons learnedSince population
size is critical andphenotyping is a key informant in GS,efficient
and mass screening methodshave been identified.Single Node Cuttings
for assessmentof tuberization under long photoperiod,or canopy
temperature and NDVI todifferentiate drought resistancetolerance
represent efficient andfeasible mass screening methods. 28.
CollaboratorsElisa MIhovilovichAwais KhanMaria CarazaDavid de
KoeyerMariela Aponte453525155Stolon Number(75 DAP)-51.8 2.7 3.5 4.3
5.2 6.0 6.8 7.6 9.0Evelyn Farfan Number of breeding linesStolon
Number(1-9) 29. Ongoing work Reference population has beenincreased
to 360 breeding lines (almost3-fold and experiments will
beperformed in replicated trialsA known gene for
photoperiodresponse on chr V, StCDF1, (CyclingDOF factor ) is being
amplified in asample of the training population set tolook for
variants toward markerassisted selectionDevelop impact pathway:
RBM. CRPLinkages 30. CollaboratorsElisa MIhovilovichAwais KhanMaria
CarazaDavid de KoeyerMariela Aponte453525155Stolon Number(75
DAP)-51.8 2.7 3.5 4.3 5.2 6.0 6.8 7.6 9.0Evelyn Farfan Number of
breeding linesStolon Number(1-9) 31. Thank you 32. Genomic
Selection estimates marker effectsacross the whole genome of a
breedingpopulation based on the prediction modeldeveloped in a
training population . A trainingpopulation is a group of
individuals (breedinglines) that are both phenotyped and genotyped
.The Breeding populationthat can be used for validation) includes
thedescendants of a Training Population or a newvariety that is
related to the training population,and is only genotyped not
phenotyped , unlessthe breeder would like to validate
thepredictionsAccura