Inferring population structure in biobank-scale genomic data

Inferring ancestry with the hierarchical soft clustering approach tangleGen

Understanding the genetic ancestry of populations is central to numerous scientific and societal fields. It contributes to a better understanding of human evolutionary history, advances personalized medicine, aids in forensic identification, and allows individuals to connect to their genealogical roots. Existing methods, such as ADMIXTURE, have significantly improved our ability to infer ancestries. However, these methods typically work with a fixed number of independent ancestral populations. As a result, they provide insight into genetic admixture, but do not include a hierarchical interpretation. In particular, the intricate ancestral population structures remain difficult to unravel. Alternative methods with a consistent inheritance structure, such as hierarchical clustering, may offer benefits in terms of interpreting the inferred ancestries. Here, we present tangleGen, a soft clustering tool that transfers the hierarchical machine learning framework Tangles, which leverages graph theoretical concepts, to the field of population genetics. The hierarchical perspective of tangleGen on the composition and structure of populations improves the interpretability of the inferred ancestral relationships. Moreover, tangleGen adds a new layer of explainability, as it allows identifying the single-nucleotide polymorphisms that are responsible for the clustering structure. We demonstrate the capabilities and benefits of tangleGen for the inference of ancestral relationships, using both simulated data and data from the 1000 Genomes Project.

Estimation of genetic admixture proportions via haplotypes

Estimation of ancestral admixture is essential for creating personal genealogies, studying human history, and conducting genome-wide association studies (GWAS). The following three primary methods exist for estimating admixture coefficients. The frequentist approach directly maximizes the binomial loglikelihood. The Bayesian approach adds a reasonable prior and samples the posterior distribution. Finally, the nonparametric approach decomposes the genotype matrix algebraically. Each approach scales successfully to datasets with a million individuals and a million single nucleotide polymorphisms (SNPs). Despite their variety, all current approaches assume independence between SNPs. To achieve independence requires performing LD (linkage disequilibrium) filtering before analysis. Unfortunately, this tactic loses valuable information and usually retains many SNPs still in LD. The present paper explores the option of explicitly incorporating haplotypes in ancestry estimation. Our program, HaploADMIXTURE, operates on adjacent SNP pairs and jointly estimates their haplotype frequencies along with admixture coefficients. This more complex strategy takes advantage of the rich information available in haplotypes and ultimately yields better admixture estimates and better clustering of real populations in curated datasets.

Genetic disease risks of under-represented founder populations in New York City

The detection of founder pathogenic variants, those observed in high frequency only in a group of individuals with increased inter-relatedness, can help improve delivery of health care for that community. We identified 16 groups with shared ancestry, based on genomic segments that are shared through identity by descent (IBD), in New York City using the genomic data of 25,366 residents from the All Of Us Research Program and the Mount Sinai BioMe biobank. From these groups we defined 8 as founder populations, mostly communities currently under-represented in medical genomics research, such as Puerto Rican, Garifuna and Filipino/Pacific Islanders. The enrichment analysis of ClinVar pathogenic or likely pathogenic (P/LP) variants in each group identified 202 of these damaging variants across the 8 founder populations. We confirmed disease-causing variants previously reported to occur at increased frequencies in Ashkenazi Jewish and Puerto Rican genetic ancestry groups, but most of the damaging variants identified have not been previously associated with any such founder populations, and most of these founder populations have not been described to have increased prevalence of the associated rare disease. Twenty-five of 51 variants meeting Tier 2 clinical screening criteria (1/100 carrier frequency within these founder groups) have never previously been reported. We show how population structure studies can provide insights into rare diseases disproportionately affecting under-represented founder populations, delivering a health care benefit but also a potential source of stigmatization of these communities, who should be part of the decision-making about implementation into health care delivery.

Decoding triancestral origins, archaic introgression, and natural selection in the Japanese population by whole-genome sequencing

We generated Japanese Encyclopedia of Whole-Genome/Exome Sequencing Library (JEWEL), a high-depth whole-genome sequencing dataset comprising 3256 individuals from across Japan. Analysis of JEWEL revealed genetic characteristics of the Japanese population that were not discernible using microarray data. First, rare variant-based analysis revealed an unprecedented fine-scale genetic structure. Together with population genetics analysis, the present-day Japanese can be decomposed into three ancestral components. Second, we identified unreported loss-of-function (LoF) variants and observed that for specific genes, LoF variants appeared to be restricted to a more limited set of transcripts than would be expected by chance, with PTPRD as a notable example. Third, we identified 44 archaic segments linked to complex traits, including a Denisovan-derived segment at NKX6-1 associated with type 2 diabetes. Most of these segments are specific to East Asians. Fourth, we identified candidate genetic loci under recent natural selection. Overall, our work provided insights into genetic characteristics of the Japanese population.

Disease risk and healthcare utilization among ancestrally diverse groups in the Los Angeles region

An individual's disease risk is affected by the populations that they belong to, due to shared genetics and environmental factors. The study of fine-scale populations in clinical care is important for identifying and reducing health disparities and for developing personalized interventions. To assess patterns of clinical diagnoses and healthcare utilization by fine-scale populations, we leveraged genetic data and electronic medical records from 35,968 patients as part of the UCLA ATLAS Community Health Initiative. We defined clusters of individuals using identity by descent, a form of genetic relatedness that utilizes shared genomic segments arising due to a common ancestor. In total, we identified 376 clusters, including clusters with patients of Afro-Caribbean, Puerto Rican, Lebanese Christian, Iranian Jewish and Gujarati ancestry. Our analysis uncovered 1,218 significant associations between disease diagnoses and clusters and 124 significant associations with specialty visits. We also examined the distribution of pathogenic alleles and found 189 significant alleles at elevated frequency in particular clusters, including many that are not regularly included in population screening efforts. Overall, this work progresses the understanding of health in understudied communities and can provide the foundation for further study into health inequities.

Neural ADMIXTURE for rapid genomic clustering

Characterizing the genetic structure of large cohorts has become increasingly important as genetic studies extend to massive, increasingly diverse biobanks. Popular methods decompose individual genomes into fractional cluster assignments with each cluster representing a vector of DNA variant frequencies. However, with rapidly increasing biobank sizes, these methods have become computationally intractable. Here we present Neural ADMIXTURE, a neural network autoencoder that follows the same modeling assumptions as the current standard algorithm, ADMIXTURE, while reducing the compute time by orders of magnitude surpassing even the fastest alternatives. One month of continuous compute using ADMIXTURE can be reduced to just hours with Neural ADMIXTURE. A multi-head approach allows Neural ADMIXTURE to offer even further acceleration by calculating multiple cluster numbers in a single run. Furthermore, the models can be stored, allowing cluster assignment to be performed on new data in linear time without needing to share the training samples.

Causal effects on complex traits are similar for common variants across segments of different continental ancestries within admixed individuals

Individuals of admixed ancestries (for example, African Americans) inherit a mosaic of ancestry segments (local ancestry) originating from multiple continental ancestral populations. This offers the unique opportunity of investigating the similarity of genetic effects on traits across ancestries within the same population. Here we introduce an approach to estimate correlation of causal genetic effects (radmix) across local ancestries and analyze 38 complex traits in African-European admixed individuals (N = 53,001) to observe very high correlations (meta-analysis radmix = 0.95, 95% credible interval 0.93-0.97), much higher than correlation of causal effects across continental ancestries. We replicate our results using regression-based methods from marginal genome-wide association study summary statistics. We also report realistic scenarios where regression-based methods yield inflated heterogeneity-by-ancestry due to ancestry-specific tagging of causal effects, and/or polygenicity. Our results motivate genetic analyses that assume minimal heterogeneity in causal effects by ancestry, with implications for the inclusion of ancestry-diverse individuals in studies.

Genomic evidence for adaptation to tuberculosis in the Andes before European contact

Most studies focusing on human high-altitude adaptation in the Andean highlands have thus far been focused on Peruvian populations. We present high-coverage whole genomes from Indigenous people living in the Ecuadorian highlands and perform multi-method scans to detect positive natural selection. We identified regions of the genome that show signals of strong selection to both cardiovascular and hypoxia pathways, which are distinct from those uncovered in Peruvian populations. However, the strongest signals of selection were related to regions of the genome that are involved in immune function related to tuberculosis. Given our estimated timing of this selection event, the Indigenous people of Ecuador may have adapted to Mycobacterium tuberculosis thousands of years before the arrival of Europeans. Furthermore, we detect a population collapse that coincides with the arrival of Europeans, which is more severe than other regions of the Andes, suggesting differing effects of contact across high-altitude populations.

Unsupervised discovery of ancestry-informative markers and genetic admixture proportions in biobank-scale datasets

Admixture estimation plays a crucial role in ancestry inference and genome-wide association studies (GWASs). Computer programs such as ADMIXTURE and STRUCTURE are commonly employed to estimate the admixture proportions of sample individuals. However, these programs can be overwhelmed by the computational burdens imposed by the 10⁵ to 10⁶ samples and millions of markers commonly found in modern biobanks. An attractive strategy is to run these programs on a set of ancestry-informative SNP markers (AIMs) that exhibit substantially different frequencies across populations. Unfortunately, existing methods for identifying AIMs require knowing ancestry labels for a subset of the sample. This supervised learning approach creates a chicken and the egg scenario. In this paper, we present an unsupervised, scalable framework that seamlessly carries out AIM selection and likelihood-based estimation of admixture proportions. Our simulated and real data examples show that this approach is scalable to modern biobank datasets. OpenADMIXTURE, our Julia implementation of the method, is open source and available for free.