polyRAD: Genotype Calling with Uncertainty from Sequencing Data in Polyploids and Diploids

Abstract

Low or uneven read depth is a common limitation of genotyping-by-sequencing (GBS) and restriction site-associated DNA sequencing (RAD-seq), resulting in high missing data rates, heterozygotes miscalled as homozygotes, and uncertainty of allele copy number in heterozygous polyploids. Bayesian genotype calling can mitigate these issues, but previously has only been implemented in software that requires a reference genome or uses priors that may be inappropriate for the population. Here we present several novel Bayesian algorithms that estimate genotype posterior probabilities, all of which are implemented in a new R package, polyRAD. Appropriate priors can be specified for mapping populations, populations in Hardy-Weinberg equilibrium, or structured populations, and in each case can be informed by genotypes at linked markers. The polyRAD software imports read depth from several existing pipelines, and outputs continuous or discrete numerical genotypes suitable for analyses such as genome-wide association and genomic prediction.

Training Population Optimization for Genomic Selection in Miscanthus

“…For the 09F2 population, each marker had to have at least 200 individuals with reads and at least 30 individuals with the minor allele. An R script using polyRAD (Clark et al 2019c) was then used for genotype calling, and a matrix of posterior mean genotypes, scaled from 0 to 1, was exported for the combined dataset. The subsequent set of 5,140 markers were used to fit and evaluate all GS models considered in the analysis of the observed traits.…”

Priors for Genotyping Polyploids

“…Here, we only analyzed one SNP at a time, though some recent approaches have also attempted to borrow strength between SNPs [Blischak et al, 2018, Clark et al, 2019]. In particular, Clark et al [2019] tries to account for linkage disequilibrium (LD) between SNPs.…”

“…Due to constraints in sequencing depth, many genotyping methods incorporate empirical Bayes techniques to borrow strength between individuals [Voorrips et al, 2011, Serang et al, 2012, Maruki and Lynch, 2017, Blischak et al, 2018, Gerard et al, 2018, Clark et al, 2019]. These methods posit some class of prior distributions for the possible genotypes, select a distribution among this class using maximum marginal likelihood, and incorporate this prior genotype distribution in a Bayesian inference scheme to derive a posterior genotype distribution from which genotype calls are made.…”

“…However, computation is a major concern as there are often tens or hundreds of thousands of SNPs to genotype in a single dataset, and even increases of seconds can make a genotyping method untenable for applied researchers. Thus, for computational reasons, genotyping methods either assume these parameters are known a priori [Maruki and Lynch, 2017, Blischak et al, 2018, Clark et al, 2019] or estimate them simultaneously with the prior genotype distribution [Voorrips et al, 2011, Gerard et al, 2018].…”

A roadmap for research in octoploid strawberry

“…Underlying computational challenges associated with genotyping by sequencing (GBS) and other nextgeneration sequencing (NGS) facilitated approaches have limited their widespread application in octoploid strawberry thus far 36,37 . The challenges are similar across species, but obviously exacerbated in allogamous polyploids: uneven and inadequate sequencing depth, copy number uncertainty, heterozygote miscalling, missing data, sequencing errors, etc., all of which challenge the integration of DNA variant information across studies [38][39][40] . As with the other DNA marker genotyping approaches reviewed here, the first GBS study in octoploid strawberry utilized the diploid F. vesca reference genome in combination with a phylogenetic approach (POLiMAPS) for aligning, classifying, and calling DNA variants 9,36 .…”