A Likelihood-Free Estimator of Population Structure Bridging Admixture Models and Principal Components Analysis

Cabreros, Irineo; Storey, J. Douglas

doi:10.1534/genetics.119.302159

Cited by 35 publications

(49 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to have a better review of the performance of Neural ADMIXTURE, we compare the time and quality performance not only against ADMIXTURE, but also against AlStructure [28] and TeraStructure [29], which were designed to improve the scalability of admixture models. TeraStructure iteratively computes Q and F while avoiding a high computational load by subsampling SNPs at every iteration, which makes the algorithm fast.…”

Section: Benchmark Setupmentioning

confidence: 99%

See 1 more Smart Citation

Neural ADMIXTURE: rapid population clustering with autoencoders

Bustamante

et al. 2021

Preprint

View full text Add to dashboard Cite

Characterizing the genetic substructure of large cohorts has become increasingly important as genetic association and prediction studies are extended to massive, increasingly diverse, biobanks. ADMIXTURE and STRUCTURE are widely used unsupervised clustering algorithms for characterizing such ancestral genetic structure. These methods decompose individual genomes into fractional cluster assignments with each cluster representing a vector of DNA marker frequencies. The assignments, and clusters, provide an interpretable representation for geneticists to describe population substructure at the sample level. However, with the rapidly increasing size of population biobanks and the growing numbers of variants genotyped (or sequenced) per sample, such traditional methods become computationally intractable. Furthermore, multiple runs with different hyperparameters are required to properly depict the population clustering using these traditional methods, increasing the computational burden. This can lead to days of compute. In this work we present Neural ADMIXTURE, a neural network autoencoder that follows the same modeling assumptions as ADMIXTURE, providing similar (or better) clustering, while reducing the compute time by orders of magnitude. In addition, this network can include multiple outputs, providing the equivalent results as running the original ADMIXTURE algorithm many times with different numbers of clusters. These models can also be stored, allowing later cluster assignment to be performed with a linear computational time.

show abstract

Section: Benchmark Setupmentioning

confidence: 99%

“…All-Chms ADMIXTURE [4,5] .042 .153 .062 >1 day -AlStructure [28] .064 .159 .032 06:04:28 -TeraStructure [29] .033 .133 -02:12:46 -HaploNet [30] . 2.…”

Section: Datasetmentioning

confidence: 99%

Neural ADMIXTURE: rapid population clustering with autoencoders

Bustamante

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The number of ancestors and ancestry coefficients were estimated using the model-based method ALStructure. (Cabreros and Storey 2019). ALStructure uses a likelihoodfree algorithm to derive estimates through minimal model assumptions and is generally superior to existing likelihood-based methods in terms of accuracy and computational speed (Cabreros and Storey 2019).…”

Section: Analysis Of Population Structurementioning

confidence: 99%

“…(Cabreros and Storey 2019). ALStructure uses a likelihoodfree algorithm to derive estimates through minimal model assumptions and is generally superior to existing likelihood-based methods in terms of accuracy and computational speed (Cabreros and Storey 2019). The method does not assume Hardy-Weinberg equilibrium within populations, but defines the number of ancestral populations (K ) as the rank of a matrix consisting of individual-specific allele frequencies (Leek 2011).…”

Section: Analysis Of Population Structurementioning

confidence: 99%

See 1 more Smart Citation

Physical geography, isolation by distance and environmental variables shape genomic variation of wild barley (Hordeum vulgare L. ssp. spontaneum) in the Southern Levant

Chang

Fridman

Mascher

et al. 2021

Preprint

View full text Add to dashboard Cite

Determining the extent of genetic variation that reflects local adaptation in crop wild relatives is of interest to discovering useful genetic diversity for plant breeding. We investigated the association of genomic variation with geographical and environmental factors in wild barley (Hordeum vulgare L. ssp. spontaneum) populations of the Southern Levant using genotyping-by-sequencing (GBS) of 244 accessions of the Barley1K+ collection. Inference of population structure resulted in four genetic clusters that corresponded to eco-geographical habitats and a significant association of lower gene flow rates with geographical barriers, e.g. the Judaean Mountains and the Sea of Galilee. Redundancy analysis (RDA) revealed that spatial autocorrelation explained 45% and environmental variables explained 15% of total genomic variation. Only 4.5% of genomic variation was exclusively attributed to environmental variation if the component confounded with spatial autocorrelation was excluded. A synthetic environmental variable combining latitude, solar radiation, and accumulated precipitation explained the highest proportion of genomic variation (3.9%). After correcting for population structure, soil water capacity was the most important environmental variable explaining 1.18% of genomic variation. Genome scans with outlier analysis and genome-environment association studies were conducted to identify signatures of adaptation. RDA and outlier methods jointly detected selection signatures in the pericentromeric regions of chromosome 3H, 4H, and 5H, but they mostly disappeared after correction for population structure. In conclusion, adaptation to the highly diverse environments of the Southern Levant over short geographical ranges has a small effect on the genomic diversity of wild barley highlighting the importance of non-selective forces in genetic differentiation.

show abstract

Portability of 245 polygenic scores when derived from the UK Biobank and applied to 9 ancestry groups from the same cohort

Privé

Aschard

Carmi

et al. 2022

The American Journal of Human Genetics

206

153

View full text Add to dashboard Cite

A Likelihood-Free Estimator of Population Structure Bridging Admixture Models and Principal Components Analysis

Abstract: Characterizing genetic variation in humans is an important task in statistical genetics, enabling disease-gene mapping in genome-wide association studies (GWAS) and informing studies of human evolutionary history. A common approach to quantifying genetic variation...

Cited by 35 publications

References 41 publications

Neural ADMIXTURE: rapid population clustering with autoencoders

Neural ADMIXTURE: rapid population clustering with autoencoders

Physical geography, isolation by distance and environmental variables shape genomic variation of wild barley (Hordeum vulgare L. ssp. spontaneum) in the Southern Levant

Portability of 245 polygenic scores when derived from the UK Biobank and applied to 9 ancestry groups from the same cohort

Contact Info

Product

Resources

About