Accurate genotypes on the Eureka™ Genotyping Solution with Bayesian clustering

The Eureka™ Genotyping Solution is an affordable, low- to mid-plex, high-throughput genotyping assay that uses common next-generation sequencing (NGS) platforms for signal readout. It enables the detection of tens to thousands of genetic markers which are increasingly in demand for routine crop and animal agrigenomics testing.

The Eureka™ NGS read counts are appropriately scaled, normalized, and transformed to process genotyping in a cluster-based framework. The BRLMM-P two-dimensional clustering algorithm, which was originally developed for genotyping single-nucleotide polymorphisms (SNPs) and insertion/deletions (indels) on mid- to high-plex microarrays, has been extended for application on NGS reads. The algorithm has successfully been implemented across panels designed to study many organisms, and it incorporates methods for accurately genotyping samples originating from normal and inbred populations.

The Bayesian step in BRLMM-P gives it the flexibility to automatically adapt to cluster patterns exhibited by diploid and various levels of allopolyploidy. It adapts pre-positioned genotype cluster locations called “priors” to the sample data and computes three posterior cluster locations. Genotype cluster locations are defined by two-dimensional means and variances for AA, AB, and BB genotype cluster distributions. Priors can be generic, meaning the same pre-positioned location is provided for every SNP, or SNP-specific, meaning different pre-positioned locations are provided on a SNP-by-SNP basis.

Eureka™ NGS read counts ratios, similar to microarray intensities, have shown to exhibit consistency across batches, resulting in repeatable spatial properties of genotype clusters. This allows SNP-specific-priors, developed under controlled environments, to generalize. The process of prior development and genotype performance improvement is discussed.