Genetic diversity analysis of highly incomplete SNP genotype data with 6 imputations : an empirical assessment

17 Genotyping by sequencing (GBS) has recently emerged as a promising genomic approach for 18 assessing genetic diversity on a genome-wide scale. However, concerns are not lacking about the 19 uniquely large unbalance in GBS genotype data. While some genotype imputation has been 20 proposed to infer missing observations, little is known about the reliability of a genetic diversity 21 analysis of GBS data with up to 90% of observations missing. Here we performed an empirical 22 assessment of accuracy in genetic diversity analysis of highly incomplete SNP genotypes with 23 imputations. Three large SNP genotype data sets for corn, wheat and rice were acquired, and 24 missing data with up to 90% of missing observations were randomly generated and then imputed 25 for missing genotypes with three map-independent imputation methods. Estimating 26 heterozygosity and inbreeding coefficient from original, missing and imputed data revealed 27 variable patterns of bias from assessed levels of missingness and genotype imputation, but the 28 estimation biases were smaller for missing data without genotype imputation. The estimates of 29 genetic differentiation were rather robust up to 90% of missing observations, but became 30 substantially biased when missing genotypes were imputed. The estimates of topology accuracy 31 for four representative samples of interested groups were generally reduced with increased levels 32 of missing genotypes. Probabilistic principal component analysis based imputation performed 33 better in terms of topology accuracy than those analyses of missing data without genotype 34 imputation. These findings are not only significant for understanding the reliability of the genetic 35 diversity analysis with respect to large missing data and genotype imputation, but also are 36 instructive for performing a proper genetic diversity analysis of highly incomplete GBS or other 37 genotype data. 38