A one penny imputed genome from next generation reference panels

Abstract

Genotype imputation is commonly performed in genome-wide association studies because it greatly increases the number of markers that can be tested for association with a trait. In general, one should perform genotype imputation using the largest reference panel that is available because the number of accurately imputed variants increases with reference panel size. However, one impediment to using larger reference panels is the increased computational cost of imputation. We present a new genotype imputation method, Beagle 5.0, which greatly reduces the computational cost of imputation from large reference panels. We compare Beagle 5.0 with Beagle 4.1, Impute4, Minimac3, and Minimac4 using 1000 Genomes Project data, Haplotype Reference Consortium data, and simulated data for 10k, 100k, 1M, and 10M reference samples. All methods produce nearly identical accuracy, but Beagle 5.0 has the lowest computation time and the best scaling of computation time with increasing reference panel size. For 10k, 100k, 1M, and 10M reference samples and 1000 phased target samples, Beagle 5.0's computation time is 3x (10k), 12x (100k), 43x (1M), and 533x (10M) faster than the fastest alternative method. Cost data from the Amazon Elastic Compute Cloud show that Beagle 5.0 can perform genomewide imputation from 10M reference samples into 1000 phased target samples at a cost of less than one US cent per sample.Beagle 5.0 is freely available from https://faculty.washington.edu/browning/beagle/beagle.html.

Genomic consequences of population decline in critically endangered pangolins and their demographic histories

“…We used the multiple sequential Markovian coalescent (MSMC) model [99] based on the phased haplotypes to estimate the divergence time and demographic history of Taiwan pangolin. The haplotypes were phased with BEAGLE v.5.0 [100]. The mutation rate (μ) and the generation time were set as the same as those described above.…”

Whole‐genome sequence analysis reveals candidate genomic footprints and genes associated with reproductive traits in Thoroughbred horse

“…Finally, to obtain high‐quality results for further analyses, loci with RMS mapping quality <25 and genotyping quality <40 were filtered using VariantFiltration command in GATK 3.4. Here, we also imputed SNPs with <10% missing frequency, and haplotype for each chromosome were deduced by BEAGLE (Version 5.0) (Browning, Zhou, & Browning, ).…”

Genotype Imputation in Genome‐Wide Association Studies

“…A number of widely‐used programs are available for genotype imputation, including MaCH/minimac, IMPUTE2, BEAGLE, PLINK, and fastPHASE, each implementing different algorithms and each with varying limitations and accuracy (Browning & Browning, ; Browning, Zhou, & Browning, ; Howie, Fuchsberger, Stephens, Marchini, & Abecasis, ; Howie, Donnelly, & Marchini, ; Li, Willer, Ding, Scheet, & Abecasis, ; Purcell et al., ; Scheet & Stephens, ). PLINK and BEAGLE are more efficient algorithms in terms of computational burden, as they utilize relatively small windows of adjacent SNPs to construct localized haplotypes for imputation.…”

“…fastPHASE can also be used for localized imputation, and both BEAGLE and fastPHASE use a Hidden Markov Model (HMM) approach in localized haplotype‐cluster models. BEAGLE however uses a more efficient approach than fastPHASE, fitting a haplotype‐cluster model using a one‐step algorithm based on current haplotype estimates without re‐estimating model parameters at each iteration, drastically reducing computation time and burden (Browning et al., ; Browning & Browning, ). BEAGLE also has a notably lower error rate compared to fastPHASE under most modeling conditions (Marchini & Howie, ).…”