A weakly structured stem for human origins in Africa.

Despite broad agreement that Homo sapiens originated in Africa, considerable uncertainty surrounds specific models of divergence and migration across the continent1. Progress is hampered by a shortage of fossil and genomic data, as well as variability in previous estimates of divergence times1. Here we seek to discriminate among such models by considering linkage disequilibrium and diversity-based statistics, optimized for rapid, complex demographic inference2. We infer detailed demographic models for populations across Africa, including eastern and western representatives, and newly sequenced whole genomes from 44 Nama (Khoe-San) individuals from southern Africa. We infer a reticulated African population history in which present-day population structure dates back to Marine Isotope Stage 5. The earliest population divergence among contemporary populations occurred 120,000 to 135,000 years ago and was preceded by links between two or more weakly differentiated ancestral Homo populations connected by gene flow over hundreds of thousands of years. Such weakly structured stem models explain patterns of polymorphism that had previously been attributed to contributions from archaic hominins in Africa2-7. In contrast to models with archaic introgression, we predict that fossil remains from coexisting ancestral populations should be genetically and morphologically similar, and that only an inferred 1-4% of genetic differentiation among contemporary human populations can be attributed to genetic drift between stem populations. We show that model misspecification explains the variation in previous estimates of divergence times, and argue that studying a range of models is key to making robust inferences about deep history.

Fast and accurate Bayesian polygenic risk modeling with variational inference.

The advent of large-scale genome-wide association studies (GWASs) has motivated the development of statistical methods for phenotype prediction with single-nucleotide polymorphism (SNP) array data. These polygenic risk score (PRS) methods use a multiple linear regression framework to infer joint effect sizes of all genetic variants on the trait. Among the subset of PRS methods that operate on GWAS summary statistics, sparse Bayesian methods have shown competitive predictive ability. However, most existing Bayesian approaches employ Markov chain Monte Carlo (MCMC) algorithms, which are computationally inefficient and do not scale favorably to higher dimensions, for posterior inference. Here, we introduce variational inference of polygenic risk scores (VIPRS), a Bayesian summary statistics-based PRS method that utilizes variational inference techniques to approximate the posterior distribution for the effect sizes. Our experiments with 36 simulation configurations and 12 real phenotypes from the UK Biobank dataset demonstrated that VIPRS is consistently competitive with the state-of-the-art in prediction accuracy while being more than twice as fast as popular MCMC-based approaches. This performance advantage is robust across a variety of genetic architectures, SNP heritabilities, and independent GWAS cohorts. In addition to its competitive accuracy on the "White British" samples, VIPRS showed improved transferability when applied to other ethnic groups, with up to 1.7-fold increase in R2 among individuals of Nigerian ancestry for low-density lipoprotein (LDL) cholesterol. To illustrate its scalability, we applied VIPRS to a dataset of 9.6 million genetic markers, which conferred further improvements in prediction accuracy for highly polygenic traits, such as height.

Lessons Learned from Bugs in Models of Human History.

Simulation plays a central role in population genomics studies. Recent years have seen rapid improvements in software efficiency that make it possible to simulate large genomic regions for many individuals sampled from large numbers of populations. As the complexity of the demographic models we study grows, however, there is an ever-increasing opportunity to introduce bugs in their implementation. Here, we describe two errors made in defining population genetic models using the msprime coalescent simulator that have found their way into the published record. We discuss how these errors have affected downstream analyses and give recommendations for software developers and users to reduce the risk of such errors.

Accounting for long-range correlations in genome-wide simulations of large cohorts.

Coalescent simulations are widely used to examine the effects of evolution and demographic history on the genetic makeup of populations. Thanks to recent progress in algorithms and data structures, simulators such as the widely-used msprime now provide genome-wide simulations for millions of individuals. However, this software relies on classic coalescent theory and its assumptions that sample sizes are small and that the region being simulated is short. Here we show that coalescent simulations of long regions of the genome exhibit large biases in identity-by-descent (IBD), long-range linkage disequilibrium (LD), and ancestry patterns, particularly when the sample size is large. We present a Wright-Fisher extension to msprime, and show that it produces more realistic distributions of IBD, LD, and ancestry proportions, while also addressing more subtle biases of the coalescent. Further, these extensions are more computationally efficient than state-of-the-art coalescent simulations when simulating long regions, including whole-genome data. For shorter regions, efficiency can be maintained via a hybrid model which simulates the recent past under the Wright-Fisher model and uses coalescent simulations in the distant past.

Assumptions about frequency-dependent architectures of complex traits bias measures of functional enrichment.

Linkage-Disequilibrium Score Regression (LDSC) is a popular framework for analyzing Genome-wide Association Studies (GWAS) summary statistics that allows for estimating single nucleotide polymorphism heritability, confounding, and functional enrichment of genetic variants with different annotations. Recent work has highlighted the influence of implicit and explicit assumptions of the model on the biological interpretation of the results. In this study, we explored a formulation of LDSC that replaces the r2 measure of LD with a recently proposed unbiased estimator of the D2 statistic. In addition to modest statistical difference across estimators, this derivation highlighted implicit and unrealistic assumptions about the relationship between allele frequency, effect size, and annotation status. We carry out a systematic comparison of alternative LDSC formulations by applying them to summary statistics from 47 GWAS traits. Our results show that commonly used models likely underestimate functional enrichment. These results highlight the importance of calibrating the LDSC model to achieve a more robust understanding of polygenic traits.

Models of archaic admixture and recent history from two-locus statistics.

We learn about population history and underlying evolutionary biology through patterns of genetic polymorphism. Many approaches to reconstruct evolutionary histories focus on a limited number of informative statistics describing distributions of allele frequencies or patterns of linkage disequilibrium. We show that many commonly used statistics are part of a broad family of two-locus moments whose expectation can be computed jointly and rapidly under a wide range of scenarios, including complex multi-population demographies with continuous migration and admixture events. A full inspection of these statistics reveals that widely used models of human history fail to predict simple patterns of linkage disequilibrium. To jointly capture the information contained in classical and novel statistics, we implemented a tractable likelihood-based inference framework for demographic history. Using this approach, we show that human evolutionary models that include archaic admixture in Africa, Asia, and Europe provide a much better description of patterns of genetic diversity across the human genome. We estimate that an unidentified, deeply diverged population admixed with modern humans within Africa both before and after the split of African and Eurasian populations, contributing 4 - 8% genetic ancestry to individuals in world-wide populations.

UMAP reveals cryptic population structure and phenotype heterogeneity in large genomic cohorts.

Human populations feature both discrete and continuous patterns of variation. Current analysis approaches struggle to jointly identify these patterns because of modelling assumptions, mathematical constraints, or numerical challenges. Here we apply uniform manifold approximation and projection (UMAP), a non-linear dimension reduction tool, to three well-studied genotype datasets and discover overlooked subpopulations within the American Hispanic population, fine-scale relationships between geography, genotypes, and phenotypes in the UK population, and cryptic structure in the Thousand Genomes Project data. This approach is well-suited to the influx of large and diverse data and opens new lines of inquiry in population-scale datasets.

Unbiased Estimation of Linkage Disequilibrium from Unphased Data.

Linkage disequilibrium (LD) is used to infer evolutionary history, to identify genomic regions under selection, and to dissect the relationship between genotype and phenotype. In each case, we require accurate estimates of LD statistics from sequencing data. Unphased data present a challenge because multilocus haplotypes cannot be inferred exactly. Widely used estimators for the common statistics r2 and D2 exhibit large and variable upward biases that complicate interpretation and comparison across cohorts. Here, we show how to find unbiased estimators for a wide range of two-locus statistics, including D2, for both single and multiple randomly mating populations. These unbiased statistics are particularly well suited to estimate effective population sizes from unlinked loci in small populations. We develop a simple inference pipeline and use it to refine estimates of recent effective population sizes of the threatened Channel Island Fox populations.

Legacy Data Confound Genomics Studies.

Recent reports have identified differences in the mutational spectra across human populations. Although some of these reports have been replicated in other cohorts, most have been reported only in the 1000 Genomes Project (1kGP) data. While investigating an intriguing putative population stratification within the Japanese population, we identified a previously unreported batch effect leading to spurious mutation calls in the 1kGP data and to the apparent population stratification. Because the 1kGP data are used extensively, we find that the batch effects also lead to incorrect imputation by leading imputation servers and a small number of suspicious GWAS associations. Lower quality data from the early phases of the 1kGP thus continue to contaminate modern studies in hidden ways. It may be time to retire or upgrade such legacy sequencing data.

A study in scarlet: MC1R as the main predictor of red hair and exemplar of the flip-flop effect.

Genetic variation in melanocortin-1 receptor (MC1R) is a known contributor to disease-free red hair in humans. Three loss-of-function single-nucleotide variants (rs1805007, rs1805008 and rs1805009) have been established as strongly correlated with red hair. The contribution of other loss-of-function MC1R variants (in particular rs1805005, rs2228479 and rs885479) and the extent to which other genetic loci are involved in red hair colour is less well understood. Here, we used the UK Biobank cohort to capture a comprehensive list of MC1R variants contributing to red hair colour. We report a correlation with red hair for both strong-effect variants (rs1805007, rs1805008 and rs1805009) and weak-effect variants (rs1805005, rs2228479 and rs885479) and show that their coefficients differ by two orders of magnitude. On the haplotype level, both strong- and weak-effect variants contribute to the red hair phenotype, but when considered individually, weak-effect variants show a reverse, negative association with red hair. The reversal of association direction in the single-variant analysis is facilitated by a distinguishing structure of MC1R, in which loss-of-function variants are never found to co-occur on the same haplotype. The other previously reported hair colour genes' variants do not substantially improve the MC1R red hair colour predictive model. Our best model for predicting red versus other hair colours yields an unparalleled area under the receiver operating characteristic of 0.96 using only MC1R variants. In summary, we present a comprehensive statistically derived characterization of the role of MC1R variants in red hair colour and offer a powerful, economical and parsimonious model that achieves unsurpassed performance.

Inferring Transmission Histories of Rare Alleles in Population-Scale Genealogies.

Learning the transmission history of alleles through a family or population plays an important role in evolutionary, demographic, and medical genetic studies. Most classical models of population genetics have attempted to do so under the assumption that the genealogy of a population is unavailable and that its idiosyncrasies can be described by a small number of parameters describing population size and mate choice dynamics. Large genetic samples have increased sensitivity to such modeling assumptions, and large-scale genealogical datasets become a useful tool to investigate realistic genealogies. However, analyses in such large datasets are often intractable using conventional methods. We present an efficient method to infer transmission paths of rare alleles through population-scale genealogies. Based on backward-time Monte Carlo simulations of genetic inheritance, we use an importance sampling scheme to dramatically speed up convergence. The approach can take advantage of available genotypes of subsets of individuals in the genealogy including haplotype structure as well as information about the mode of inheritance and general prevalence of a mutation or disease in the population. Using a high-quality genealogical dataset of more than three million married individuals in the Quebec founder population, we apply the method to reconstruct the transmission history of chronic atrial and intestinal dysrhythmia (CAID), a rare recessive disease. We identify the most likely early carriers of the mutation and geographically map the expected carrier rate in the present-day French-Canadian population of Quebec.

A review of UMAP in population genetics.

Uniform manifold approximation and projection (UMAP) has been rapidly adopted by the population genetics community to study population structure. It has become common in visualizing the ancestral composition of human genetic datasets, as well as searching for unique clusters of data, and for identifying geographic patterns. Here we give an overview of applications of UMAP in population genetics, provide recommendations for best practices, and offer insights on optimal uses for the technique.

Estimating the mutation load in human genomes.

Next-generation sequencing technology has facilitated the discovery of millions of genetic variants in human genomes. A sizeable fraction of these variants are predicted to be deleterious. Here, we review the pattern of deleterious alleles as ascertained in genome sequencing data sets and ask whether human populations differ in their predicted burden of deleterious alleles - a phenomenon known as mutation load. We discuss three demographic models that are predicted to affect mutation load and relate these models to the evidence (or the lack thereof) for variation in the efficacy of purifying selection in diverse human genomes. We also emphasize why accurate estimation of mutation load depends on assumptions regarding the distribution of dominance and selection coefficients - quantities that remain poorly characterized for current genomic data sets.

Human Demographic History Impacts Genetic Risk Prediction across Diverse Populations.

The vast majority of genome-wide association studies (GWASs) are performed in Europeans, and their transferability to other populations is dependent on many factors (e.g., linkage disequilibrium, allele frequencies, genetic architecture). As medical genomics studies become increasingly large and diverse, gaining insights into population history and consequently the transferability of disease risk measurement is critical. Here, we disentangle recent population history in the widely used 1000 Genomes Project reference panel, with an emphasis on populations underrepresented in medical studies. To examine the transferability of single-ancestry GWASs, we used published summary statistics to calculate polygenic risk scores for eight well-studied phenotypes. We identify directional inconsistencies in all scores; for example, height is predicted to decrease with genetic distance from Europeans, despite robust anthropological evidence that West Africans are as tall as Europeans on average. To gain deeper quantitative insights into GWAS transferability, we developed a complex trait coalescent-based simulation framework considering effects of polygenicity, causal allele frequency divergence, and heritability. As expected, correlations between true and inferred risk are typically highest in the population from which summary statistics were derived. We demonstrate that scores inferred from European GWASs are biased by genetic drift in other populations even when choosing the same causal variants and that biases in any direction are possible and unpredictable. This work cautions that summarizing findings from large-scale GWASs may have limited portability to other populations using standard approaches and highlights the need for generalized risk prediction methods and the inclusion of more diverse individuals in medical genomics.

Alternative polyadenylation confers Pten mRNAs stability and resistance to microRNAs.

Fine regulation of the phosphatase and tensin homologue (PTEN) phosphatase dosage is critical for homeostasis and tumour suppression. The 3'-untranslated region (3'-UTR) of Pten mRNA was extensively linked to post-transcriptional regulation by microRNAs (miRNAs). In spite of this critical regulatory role, alternative 3'-UTRs of Pten have not been systematically characterized. Here, we reveal an important diversity of Pten mRNA isoforms generated by alternative polyadenylation sites. Several 3'-UTRs are co-expressed and their relative expression is dynamically regulated. In spite of encoding multiple validated miRNA-binding sites, longer isoforms are largely refractory to miRNA-mediated silencing, are more stable and contribute to the bulk of PTEN protein and signalling functions. Taken together, our results warrant a mechanistic re-interpretation of the post-transcriptional mechanisms involving Pten mRNAs and raise concerns on how miRNA-binding sites are being validated.

Intratumor Heterogeneity and Circulating Tumor Cell Clusters.

Genetic diversity plays a central role in tumor progression, metastasis, and resistance to treatment. Experiments are shedding light on this diversity at ever finer scales, but interpretation is challenging. Using recent progress in numerical models, we simulate macroscopic tumors to investigate the interplay between growth dynamics, microscopic composition, and circulating tumor cell cluster diversity. We find that modest differences in growth parameters can profoundly change microscopic diversity. Simple outwards expansion leads to spatially segregated clones and low diversity, as expected. However, a modest cell turnover can result in an increased number of divisions and mixing among clones resulting in increased microscopic diversity in the tumor core. Using simulations to estimate power to detect such spatial trends, we find that multiregion sequencing data from contemporary studies is marginally powered to detect the predicted effects. Slightly larger samples, improved detection of rare variants, or sequencing of smaller biopsies or circulating tumor cell clusters would allow one to distinguish between leading models of tumor evolution. The genetic composition of circulating tumor cell clusters, which can be obtained from non-invasive blood draws, is therefore informative about tumor evolution and its metastatic potential.

nanoCAGE reveals 5' UTR features that define specific modes of translation of functionally related MTOR-sensitive mRNAs.

The diversity of MTOR-regulated mRNA translation remains unresolved. Whereas ribosome-profiling suggested that MTOR almost exclusively stimulates translation of the TOP (terminal oligopyrimidine motif) and TOP-like mRNAs, polysome-profiling indicated that MTOR also modulates translation of mRNAs without the 5' TOP motif (non-TOP mRNAs). We demonstrate that in ribosome-profiling studies, detection of MTOR-dependent changes in non-TOP mRNA translation was obscured by low sensitivity and methodology biases. Transcription start site profiling using nano-cap analysis of gene expression (nanoCAGE) revealed that not only do many MTOR-sensitive mRNAs lack the 5' TOP motif but that 5' UTR features distinguish two functionally and translationally distinct subsets of MTOR-sensitive mRNAs: (1) mRNAs with short 5' UTRs enriched for mitochondrial functions, which require EIF4E but are less EIF4A1-sensitive; and (2) long 5' UTR mRNAs encoding proliferation- and survival-promoting proteins, which are both EIF4E- and EIF4A1-sensitive. Selective inhibition of translation of mRNAs harboring long 5' UTRs via EIF4A1 suppression leads to sustained expression of proteins involved in respiration but concomitant loss of those protecting mitochondrial structural integrity, resulting in apoptosis. Conversely, simultaneous suppression of translation of both long and short 5' UTR mRNAs by MTOR inhibitors results in metabolic dormancy and a predominantly cytostatic effect. Thus, 5' UTR features define different modes of MTOR-sensitive translation of functionally distinct subsets of mRNAs, which may explain the diverse impact of MTOR and EIF4A inhibitors on neoplastic cells.

The Great Migration and African-American Genomic Diversity.

We present a comprehensive assessment of genomic diversity in the African-American population by studying three genotyped cohorts comprising 3,726 African-Americans from across the United States that provide a representative description of the population across all US states and socioeconomic status. An estimated 82.1% of ancestors to African-Americans lived in Africa prior to the advent of transatlantic travel, 16.7% in Europe, and 1.2% in the Americas, with increased African ancestry in the southern United States compared to the North and West. Combining demographic models of ancestry and those of relatedness suggests that admixture occurred predominantly in the South prior to the Civil War and that ancestry-biased migration is responsible for regional differences in ancestry. We find that recent migrations also caused a strong increase in genetic relatedness among geographically distant African-Americans. Long-range relatedness among African-Americans and between African-Americans and European-Americans thus track north- and west-bound migration routes followed during the Great Migration of the twentieth century. By contrast, short-range relatedness patterns suggest comparable mobility of ∼15-16km per generation for African-Americans and European-Americans, as estimated using a novel analytical model of isolation-by-distance.