P0502An Improved Genotyping Workflow for Genomic Selection Using the Axiom® Genotyping Solution from Affymetrix
An Improved Genotyping Workflow for Genomic Selection Using the Axiom® Genotyping Solution from Affymetrix
Background:
The availability of highly parallel SNP genotyping technologies at low cost has enabled Genomic Selection (GS) as a practical tool for breeders that replaced traditional Marker-Assisted Selection.
GS requires highly accurate and reproducible genotype calls often from small batches of samples.
Methods:
The standard analysis pipeline recommended by Affymetrix (Axiom® Best Practice Workflow), which uses AxiomGT1 as a genotyping algorithm, usually requires a larger collection of input samples to maximize accuracy and reproducibility than would be applicable in GS.
A new analytical workflow has been developed, called “Axiom® Genomic Selection workflow” (GSW), and is targeted for the needs of the breeders. The new workflow achieves high accuracy and reproducibility using small sample data sets. Axiom® GSW exploits the Bayesian nature of the AxiomGT1 algorithm to use as “prior” information a model of the expected cluster positions and variances for each SNP. By adopting SNP Specific priors, the algorithm has an expectation of the cluster locations (therefore increasing reproducibility), but retains the crucial ability to update the priors to account for plate specific variations. 384 bovine samples were processed on a custom-design Axiom® array in parallel. Array data were genotyped using the Axiom® GSW method. SNP cluster models were empirically determined and tested
Result:
The new workflow was validated on a bovine dataset of 5376 samples. A significant improvement in the batch-to-batch reproducibility was observed. The concordance rate increased from 98% to greater than 99.7%. Marginal improvements on average Call Rates and number of recommended SNPs were also observed.
The availability of highly parallel SNP genotyping technologies at low cost has enabled Genomic Selection (GS) as a practical tool for breeders that replaced traditional Marker-Assisted Selection.
GS requires highly accurate and reproducible genotype calls often from small batches of samples.
Methods:
The standard analysis pipeline recommended by Affymetrix (Axiom® Best Practice Workflow), which uses AxiomGT1 as a genotyping algorithm, usually requires a larger collection of input samples to maximize accuracy and reproducibility than would be applicable in GS.
A new analytical workflow has been developed, called “Axiom® Genomic Selection workflow” (GSW), and is targeted for the needs of the breeders. The new workflow achieves high accuracy and reproducibility using small sample data sets. Axiom® GSW exploits the Bayesian nature of the AxiomGT1 algorithm to use as “prior” information a model of the expected cluster positions and variances for each SNP. By adopting SNP Specific priors, the algorithm has an expectation of the cluster locations (therefore increasing reproducibility), but retains the crucial ability to update the priors to account for plate specific variations. 384 bovine samples were processed on a custom-design Axiom® array in parallel. Array data were genotyped using the Axiom® GSW method. SNP cluster models were empirically determined and tested
Result:
The new workflow was validated on a bovine dataset of 5376 samples. A significant improvement in the batch-to-batch reproducibility was observed. The concordance rate increased from 98% to greater than 99.7%. Marginal improvements on average Call Rates and number of recommended SNPs were also observed.