Search

Single-cell technologies have transformed our understanding of human tissues. Yet, studies typically capture only a limited number of donors and disagree on cell type definitions. Integrating many single-cell datasets can address these limitations of individual studies and capture the variability present in the population. Here we present the integrated Human Lung Cell Atlas (HLCA), combining 49 datasets of the human respiratory system into a single atlas spanning over 2.4 million cells from 486 individuals. The HLCA presents a consensus cell type re-annotation with matching marker genes, including annotations of rare and previously undescribed cell types. Leveraging the number and diversity of individuals in the HLCA, we identify gene modules that are associated with demographic covariates such as age, sex and body mass index, as well as gene modules changing expression along the proximal-to-distal axis of the bronchial tree. Mapping new data to the HLCA enables rapid data annotation and interpretation. Using the HLCA as a reference for the study of disease, we identify shared cell states across multiple lung diseases, including SPP1+ profibrotic monocyte-derived macrophages in COVID-19, pulmonary fibrosis and lung carcinoma. Overall, the HLCA serves as an example for the development and use of large-scale, cross-dataset organ atlases within the Human Cell Atlas.

The inference of genome rearrangement events has been extensively studied, as they play a major role in molecular evolution. However, probabilistic evolutionary models that explicitly imitate the evolutionary dynamics of such events, as well as methods to infer model parameters, are yet to be fully utilized. Here, we developed a probabilistic approach to infer genome rearrangement rate parameters using an Approximate Bayesian Computation (ABC) framework. We developed two genome rearrangement models, a basic model, which accounts for genomic changes in gene order, and a more sophisticated one which also accounts for changes in chromosome number. We characterized the ABC inference accuracy using simulations and applied our methodology to both prokaryotic and eukaryotic empirical datasets. Knowledge of genome-rearrangement rates can help elucidate their role in evolution as well as help simulate genomes with evolutionary dynamics that reflect empirical genomes.

The Human Developmental Cell Atlas (HDCA) initiative, which is part of the Human Cell Atlas, aims to create a comprehensive reference map of cells during development. This will be critical to understanding normal organogenesis, the effect of mutations, environmental factors and infectious agents on human development, congenital and childhood disorders, and the cellular basis of ageing, cancer and regenerative medicine. Here we outline the HDCA initiative and the challenges of mapping and modelling human development using state-of-the-art technologies to create a reference atlas across gestation. Similar to the Human Genome Project, the HDCA will integrate the output from a growing community of scientists who are mapping human development into a unified atlas. We describe the early milestones that have been achieved and the use of human stem-cell-derived cultures, organoids and animal models to inform the HDCA, especially for prenatal tissues that are hard to acquire. Finally, we provide a roadmap towards a complete atlas of human development.

Crucial transitions in cancer-including tumor initiation, local expansion, metastasis, and therapeutic resistance-involve complex interactions between cells within the dynamic tumor ecosystem. Transformative single-cell genomics technologies and spatial multiplex in situ methods now provide an opportunity to interrogate this complexity at unprecedented resolution. The Human Tumor Atlas Network (HTAN), part of the National Cancer Institute (NCI) Cancer Moonshot Initiative, will establish a clinical, experimental, computational, and organizational framework to generate informative and accessible three-dimensional atlases of cancer transitions for a diverse set of tumor types. This effort complements both ongoing efforts to map healthy organs and previous large-scale cancer genomics approaches focused on bulk sequencing at a single point in time. Generating single-cell, multiparametric, longitudinal atlases and integrating them with clinical outcomes should help identify novel predictive biomarkers and features as well as therapeutically relevant cell types, cell states, and cellular interactions across transitions. The resulting tumor atlases should have a profound impact on our understanding of cancer biology and have the potential to improve cancer detection, prevention, and therapeutic discovery for better precision-medicine treatments of cancer patients and those at risk for cancer.

Chromosomal instability is a hallmark of cancer that results from ongoing errors in chromosome segregation during mitosis. Although chromosomal instability is a major driver of tumour evolution, its role in metastasis has not been established. Here we show that chromosomal instability promotes metastasis by sustaining a tumour cell-autonomous response to cytosolic DNA. Errors in chromosome segregation create a preponderance of micronuclei whose rupture spills genomic DNA into the cytosol. This leads to the activation of the cGAS-STING (cyclic GMP-AMP synthase-stimulator of interferon genes) cytosolic DNA-sensing pathway and downstream noncanonical NF-κB signalling. Genetic suppression of chromosomal instability markedly delays metastasis even in highly aneuploid tumour models, whereas continuous chromosome segregation errors promote cellular invasion and metastasis in a STING-dependent manner. By subverting lethal epithelial responses to cytosolic DNA, chromosomally unstable tumour cells co-opt chronic activation of innate immune pathways to spread to distant organs.

The recent advent of methods for high-throughput single-cell molecular profiling has catalyzed a growing sense in the scientific community that the time is ripe to complete the 150-year-old effort to identify all cell types in the human body. The Human Cell Atlas Project is an international collaborative effort that aims to define all human cell types in terms of distinctive molecular profiles (such as gene expression profiles) and to connect this information with classical cellular descriptions (such as location and morphology). An open comprehensive reference map of the molecular state of cells in healthy human tissues would propel the systematic study of physiological states, developmental trajectories, regulatory circuitry and interactions of cells, and also provide a framework for understanding cellular dysregulation in human disease. Here we describe the idea, its potential utility, early proofs-of-concept, and some design considerations for the Human Cell Atlas, including a commitment to open data, code, and community.

Background & aimsCrohn's disease (CD) has the highest prevalence in Ashkenazi Jewish populations. We sought to identify rare, CD-associated frameshift variants of high functional and statistical effects.MethodsWe performed exome sequencing and array-based genotype analyses of 1477 Ashkenazi Jewish individuals with CD and 2614 Ashkenazi Jewish individuals without CD (controls). To validate our findings, we performed genotype analyses of an additional 1515 CD cases and 7052 controls for frameshift mutations in the colony-stimulating factor 2-receptor β common subunit gene (CSF2RB). Intestinal tissues and blood samples were collected from patients with CD; lamina propria leukocytes were isolated and expression of CSF2RB and granulocyte-macrophage colony-stimulating factor-responsive cells were defined by adenomatous polyposis coli (APC) time-of-flight mass cytometry (CyTOF analysis). Variants of CSF2RB were transfected into HEK293 cells and the expression and functions of gene products were compared.ResultsIn the discovery cohort, we associated CD with a frameshift mutation in CSF2RB (P = 8.52 × 10(-4)); the finding was validated in the replication cohort (combined P = 3.42 × 10(-6)). Incubation of intestinal lamina propria leukocytes with granulocyte-macrophage colony-stimulating factor resulted in high levels of phosphorylation of signal transducer and activator of transcription (STAT5) and lesser increases in phosphorylation of extracellular signal-regulated kinase and AK straining transforming (AKT). Cells co-transfected with full-length and mutant forms of CSF2RB had reduced pSTAT5 after stimulation with granulocyte-macrophage colony-stimulating factor, compared with cells transfected with control CSF2RB, indicating a dominant-negative effect of the mutant gene. Monocytes from patients with CD who were heterozygous for the frameshift mutation (6% of CD cases analyzed) had reduced responses to granulocyte-macrophage colony-stimulating factor and markedly decreased activity of aldehyde dehydrogenase; activity of this enzyme has been associated with immune tolerance.ConclusionsIn a genetic analysis of Ashkenazi Jewish individuals, we associated CD with a frameshift mutation in CSF2RB. Intestinal monocytes from carriers of this mutation had reduced responses to granulocyte-macrophage colony-stimulating factor, providing an additional mechanism for alterations to the innate immune response in individuals with CD.

The rate at which human genomes mutate is a central biological parameter that has many implications for our ability to understand demographic and evolutionary phenomena. We present a method for inferring mutation and gene-conversion rates by using the number of sequence differences observed in identical-by-descent (IBD) segments together with a reconstructed model of recent population-size history. This approach is robust to, and can quantify, the presence of substantial genotyping error, as validated in coalescent simulations. We applied the method to 498 trio-phased sequenced Dutch individuals and inferred a point mutation rate of 1.66 × 10(-8) per base per generation and a rate of 1.26 × 10(-9) for

Although a few hundred single nucleotide polymorphisms (SNPs) suffice to infer close familial relationships, high density genome-wide SNP data make possible the inference of more distant relationships such as 2(nd) to 9(th) cousinships. In order to characterize the relationship between genetic similarity and degree of kinship given a timeframe of 100-300 years, we analyzed the sharing of DNA inferred to be identical by descent (IBD) in a subset of individuals from the 23andMe customer database (n = 22,757) and from the Human Genome Diversity Panel (HGDP-CEPH, n = 952). With data from 121 populations, we show that the average amount of DNA shared IBD in most ethnolinguistically-defined populations, for example Native American groups, Finns and Ashkenazi Jews, differs from continentally-defined populations by several orders of magnitude. Via extensive pedigree-based simulations, we determined bounds for predicted degrees of relationship given the amount of genomic IBD sharing in both endogamous and 'unrelated' population samples. Using these bounds as a guide, we detected tens of thousands of 2(nd) to 9(th) degree cousin pairs within a heterogenous set of 5,000 Europeans. The ubiquity of distant relatives, detected via IBD segments, in both ethnolinguistic populations and in large 'unrelated' populations samples has important implications for genetic genealogy, forensics and genotype/phenotype mapping studies.

[This corrects the article DOI: 10.1371/journal.pgen.1009015.].

Evidence from both GWAS and clinical observation has suggested that certain psychiatric, metabolic, and autoimmune diseases are heterogeneous, comprising multiple subtypes with distinct genomic etiologies and Polygenic Risk Scores (PRS). However, the presence of subtypes within many phenotypes is frequently unknown. We present CLiP (Correlated Liability Predictors), a method to detect heterogeneity in single GWAS cohorts. CLiP calculates a weighted sum of correlations between SNPs contributing to a PRS on the case/control liability scale. We demonstrate mathematically and through simulation that among i.i.d. homogeneous cases generated by a liability threshold model, significant anti-correlations are expected between otherwise independent predictors due to ascertainment on the hidden liability score. In the presence of heterogeneity from distinct etiologies, confounding by covariates, or mislabeling, these correlation patterns are altered predictably. We further extend our method to two additional association study designs: CLiP-X for quantitative predictors in applications such as transcriptome-wide association, and CLiP-Y for quantitative phenotypes, where there is no clear distinction between cases and controls. Through simulations, we demonstrate that CLiP and its extensions reliably distinguish between homogeneous and heterogeneous cohorts when the PRS explains as low as 3% of variance on the liability scale and cohorts comprise 50, 000 - 100, 000 samples, an increasingly practical size for modern GWAS. We apply CLiP to heterogeneity detection in schizophrenia cohorts totaling > 50, 000 cases and controls collected by the Psychiatric Genomics Consortium. We observe significant heterogeneity in mega-analysis of the combined PGC data (p-value 8.54 × 0-4), as well as in individual cohorts meta-analyzed using Fisher's method (p-value 0.03), based on significantly associated variants. We also apply CLiP-Y to detect heterogeneity in neuroticism in over 10, 000 individuals from the UK Biobank and detect heterogeneity with a p-value of 1.68 × 10-9. Scores were not significantly reduced when partitioning by known subclusters ("Depression" and "Worry"), suggesting that these factors are not the primary source of observed heterogeneity.

Genome-wide association studies (GWAS) have identified hundreds of SNPs responsible for variation in human quantitative traits. However, genome-wide-significant associations often fail to replicate across independent cohorts, in apparent inconsistency with their apparent strong effects in discovery cohorts. This limited success of replication raises pervasive questions about the utility of the GWAS field. We identify all 332 studies of quantitative traits from the NHGRI-EBI GWAS Database with attempted replication. We find that the majority of studies provide insufficient data to evaluate replication rates. The remaining papers replicate significantly worse than expected (p < 10-14), even when adjusting for regression-to-the-mean of effect size between discovery- and replication-cohorts termed the Winner's Curse (p < 10-16). We show this is due in part to misreporting replication cohort-size as a maximum number, rather than per-locus one. In 39 studies accurately reporting per-locus cohort-size for attempted replication of 707 loci in samples with similar ancestry, replication rate matched expectation (predicted 458, observed 457, p = 0.94). In contrast, ancestry differences between replication and discovery (13 studies, 385 loci) cause the most highly-powered decile of loci to replicate worse than expected, due to difference in linkage disequilibrium.

The Ashkenazi Jewish (AJ) population is important in genetics due to its high rate of Mendelian disorders. AJ appeared in Europe in the 10th century, and their ancestry is thought to comprise European (EU) and Middle-Eastern (ME) components. However, both the time and place of admixture are subject to debate. Here, we attempt to characterize the AJ admixture history using a careful application of new and existing methods on a large AJ sample. Our main approach was based on local ancestry inference, in which we first classified each AJ genomic segment as EU or ME, and then compared allele frequencies along the EU segments to those of different EU populations. The contribution of each EU source was also estimated using GLOBETROTTER and haplotype sharing. The time of admixture was inferred based on multiple statistics, including ME segment lengths, the total EU ancestry per chromosome, and the correlation of ancestries along the chromosome. The major source of EU ancestry in AJ was found to be Southern Europe (≈60-80% of EU ancestry), with the rest being likely Eastern European. The inferred admixture time was ≈30 generations ago, but multiple lines of evidence suggest that it represents an average over two or more events, pre- and post-dating the founder event experienced by AJ in late medieval times. The time of the pre-bottleneck admixture event, which was likely Southern European, was estimated to ≈25-50 generations ago.

Missing data are an unavoidable component of modern statistical genetics. Different array or sequencing technologies cover different single nucleotide polymorphisms (SNPs), leading to a complicated mosaic pattern of missingness where both individual genotypes and entire SNPs are sporadically absent. Such missing data patterns cannot be ignored without introducing bias, yet cannot be inferred exclusively from nonmissing data. In genome-wide association studies, the accepted solution to missingness is to impute missing data using external reference haplotypes. The resulting probabilistic genotypes may be analyzed in the place of genotype calls. A general-purpose paradigm, called Multiple Imputation (MI), is known to model uncertainty in many contexts, yet it is not widely used in association studies. Here, we undertake a systematic evaluation of existing imputed data analysis methods and MI. We characterize biases related to uncertainty in association studies, and find that bias is introduced both at the imputation level, when imputation algorithms generate inconsistent genotype probabilities, and at the association level, when analysis methods inadequately model genotype uncertainty. We find that MI performs at least as well as existing methods or in some cases much better, and provides a straightforward paradigm for adapting existing genotype association methods to uncertain data.

Abstract The genetic mapping of drug-response traits is often characterised by a poor signal-to-noise ratio that is placebo related and which distinguishes pharmacogenetic association studies from classical case-control studies for disease susceptibility. The goal of this study was to evaluate the statistical power of candidate gene association studies under different pharmacogenetic scenarios, with special emphasis on the placebo effect. Genotype/phenotype data were simulated, mimicking samples from clinical trials, and response to the drug was modelled as a binary trait. Association was evaluated by a logistic regression model. Statistical power was estimated as a function of the number of single nucleotide polymorphisms (SNPs) genotyped, the frequency of the placebo 'response', the genotype relative risk (GRR) of the response polymorphism, the strategy for selecting SNPs for genotyping, the number of individuals in the trial and the ratio of placebo-treated to drugtreated patients. We show that: (i) the placebo 'response' strongly affects the statistical power of association studies -- even a highly penetrant drug-response allele requires at least a 500-patient trial in order to reach 80 per cent power, several-fold more than the value estimated by standard tools that are not calibrated to pharmacogenetics; (ii) the power of a pharmacogenetic association study depends primarily on the penetrance of the response genotype and, when this penetrance is fixed, power decreases for larger placebo effects; (iii) power is dramatically increased when adding markers; (iv) an optimal study design includes a similar number of placebo- and drugtreated patients; and (v) in this setting, straightforward haplotype analysis does not seem to have an advantage over single marker analysis.

Associations between the level of single transcripts and single corresponding genetic variants, expression single nucleotide polymorphisms (eSNPs), have been extensively studied and reported. However, most expression traits are complex, involving the cooperative action of multiple SNPs at different loci affecting multiple genes. Finding these cooperating eSNPs by exhaustive search has proven to be statistically challenging. In this paper we utilized availability of sequencing data with transcriptional profiles in the same cohorts to identify two kinds of usual suspects: eSNPs that alter coding sequences or eSNPs within the span of transcription factors (TFs). We utilize a computational framework for considering triplets, each comprised of a SNP and two associated genes. We examine pairs of triplets with such cooperating source eSNPs that are both associated with the same pair of target genes. We characterize such quartets through their genomic, topological and functional properties. We establish that this regulatory structure of cooperating quartets is frequent in real data, but is rarely observed in permutations. eSNP sources are mostly located on different chromosomes and away from their targets. In the majority of quartets, SNPs affect the expression of the two gene targets independently of one another, suggesting a mutually independent rather than a directionally dependent effect. Furthermore, the directions in which the minor allele count of the SNP affects gene expression within quartets are consistent, so that the two source eSNPs either both have the same effect on the target genes or both affect one gene in the opposite direction to the other. Same-effect eSNPs are observed more often than expected by chance. Cooperating quartets reported here in a human system might correspond to bi-fans, a known network motif of four nodes previously described in model organisms. Overall, our analysis offers insights regarding the fine motif structure of human regulatory networks.

Unraveling the molecular processes that lead from genotype to phenotype is crucial for the understanding and effective treatment of genetic diseases. Knowledge of the causative genetic defect most often does not enable treatment; therefore, causal intermediates between genotype and phenotype constitute valuable candidates for molecular intervention points that can be therapeutically targeted. Mapping genetic determinants of gene expression levels (also known as expression quantitative trait loci or eQTL studies) is frequently used for this purpose, yet distinguishing causation from correlation remains a significant challenge. Here, we address this challenge using extensive, multi-environment gene expression and fitness profiling of hundreds of genetically diverse yeast strains, in order to identify truly causal intermediate genes that condition fitness in a given environment. Using functional genomics assays, we show that the predictive power of eQTL studies for inferring causal intermediate genes is poor unless performed across multiple environments. Surprisingly, although the effects of genotype on fitness depended strongly on environment, causal intermediates could be most reliably predicted from genetic effects on expression present in all environments. Our results indicate a mechanism explaining this apparent paradox, whereby immediate molecular consequences of genetic variation are shared across environments, and environment-dependent phenotypic effects result from downstream integration of environmental signals. We developed a statistical model to predict causal intermediates that leverages this insight, yielding over 400 transcripts, for the majority of which we experimentally validated their role in conditioning fitness. Our findings have implications for the design and analysis of clinical omics studies aimed at discovering personalized targets for molecular intervention, suggesting that inferring causation in a single cellular context can benefit from molecular profiling in multiple contexts.

Telomeres protect the chromosome ends from degradation and play crucial roles in cellular aging and disease. Recent studies have additionally found a correlation between psychological stress, telomere length, and health outcome in humans. However, studies have not yet explored the causal relationship between stress and telomere length, or the molecular mechanisms underlying that relationship. Using yeast as a model organism, we show that stresses may have very different outcomes: alcohol and acetic acid elongate telomeres, whereas caffeine and high temperatures shorten telomeres. Additional treatments, such as oxidative stress, show no effect. By combining genome-wide expression measurements with a systematic genetic screen, we identify the Rap1/Rif1 pathway as the central mediator of the telomeric response to environmental signals. These results demonstrate that telomere length can be manipulated, and that a carefully regulated homeostasis may become markedly deregulated in opposing directions in response to different environmental cues.

Crohn's disease (CD) is a complex disorder resulting from the interaction of intestinal microbiota with the host immune system in genetically susceptible individuals. The largest meta-analysis of genome-wide association to date identified 71 CD-susceptibility loci in individuals of European ancestry. An important epidemiological feature of CD is that it is 2-4 times more prevalent among individuals of Ashkenazi Jewish (AJ) descent compared to non-Jewish Europeans (NJ). To explore genetic variation associated with CD in AJs, we conducted a genome-wide association study (GWAS) by combining raw genotype data across 10 AJ cohorts consisting of 907 cases and 2,345 controls in the discovery stage, followed up by a replication study in 971 cases and 2,124 controls. We confirmed genome-wide significant associations of 9 known CD loci in AJs and replicated 3 additional loci with strong signal (p

As the interface between a microbe and its environment, the bacterial cell envelope has broad biological and clinical significance. While numerous biosynthesis genes and pathways have been identified and studied in isolation, how these intersect functionally to ensure envelope integrity during adaptive responses to environmental challenge remains unclear. To this end, we performed high-density synthetic genetic screens to generate quantitative functional association maps encompassing virtually the entire cell envelope biosynthetic machinery of Escherichia coli under both auxotrophic (rich medium) and prototrophic (minimal medium) culture conditions. The differential patterns of genetic interactions detected among > 235,000 digenic mutant combinations tested reveal unexpected condition-specific functional crosstalk and genetic backup mechanisms that ensure stress-resistant envelope assembly and maintenance. These networks also provide insights into the global systems connectivity and dynamic functional reorganization of a universal bacterial structure that is both broadly conserved among eubacteria (including pathogens) and an important target.

Cancer is a disease driven by a combination of inherited risk alleles coupled with the acquisition of somatic mutations, including amplification and deletion of genomic DNA. Potential relationships between the inherited and somatic aspects of the disease have only rarely been examined on a genome-wide level. Applying a novel integrative analysis of SNP and copy number measurements, we queried the tumor and normal-tissue genomes of 178 glioblastoma patients from the Cancer Genome Atlas project for preferentially amplified alleles, under the hypothesis that oncogenic germline variants will be selectively amplified in the tumor environment. Selected alleles are revealed by allelic imbalance in amplification across samples. This general approach is based on genetic principles and provides a method for identifying important tumor-related alleles. We find that SNP alleles that are most significantly overrepresented in amplicons tend to occur in genes involved with regulation of kinase and transferase activity, and many of these genes are known contributors to gliomagenesis. The analysis also implicates variants in synapse genes. By incorporating gene expression data, we demonstrate synergy between preferential allelic amplification and expression in DOCK4 and EGFR. Our results support the notion that combining germline and tumor genetic data can identify regions relevant to cancer biology.

Despite the recent rapid growth in genome-wide data, much of human variation remains entirely unexplained. A significant challenge in the pursuit of the genetic basis for variation in common human traits is the efficient, coordinated collection of genotype and phenotype data. We have developed a novel research framework that facilitates the parallel study of a wide assortment of traits within a single cohort. The approach takes advantage of the interactivity of the Web both to gather data and to present genetic information to research participants, while taking care to correct for the population structure inherent to this study design. Here we report initial results from a participant-driven study of 22 traits. Replications of associations (in the genes OCA2, HERC2, SLC45A2, SLC24A4, IRF4, TYR, TYRP1, ASIP, and MC1R) for hair color, eye color, and freckling validate the Web-based, self-reporting paradigm. The identification of novel associations for hair morphology (rs17646946, near TCHH; rs7349332, near WNT10A; and rs1556547, near OFCC1), freckling (rs2153271, in BNC2), the ability to smell the methanethiol produced after eating asparagus (rs4481887, near OR2M7), and photic sneeze reflex (rs10427255, near ZEB2, and rs11856995, near NR2F2) illustrates the power of the approach.

It has been argued that the limited genetic diversity and reduced allelic heterogeneity observed in isolated founder populations facilitates discovery of loci contributing to both Mendelian and complex disease. A strong founder effect, severe isolation, and substantial inbreeding have dramatically reduced genetic diversity in natives from the island of Kosrae, Federated States of Micronesia, who exhibit a high prevalence of obesity and other metabolic disorders. We hypothesized that genetic drift and possibly natural selection on Kosrae might have increased the frequency of previously rare genetic variants with relatively large effects, making these alleles readily detectable in genome-wide association analysis. However, mapping in large, inbred cohorts introduces analytic challenges, as extensive relatedness between subjects violates the assumptions of independence upon which traditional association test statistics are based. We performed genome-wide association analysis for 15 quantitative traits in 2,906 members of the Kosrae population, using novel approaches to manage the extreme relatedness in the sample. As positive controls, we observe association to known loci for plasma cholesterol, triglycerides, and C-reactive protein and to a compelling candidate loci for thyroid stimulating hormone and fasting plasma glucose. We show that our study is well powered to detect common alleles explaining >/=5% phenotypic variance. However, no such large effects were observed with genome-wide significance, arguing that even in such a severely inbred population, common alleles typically have modest effects. Finally, we show that a majority of common variants discovered in Caucasians have indistinguishable effect sizes on Kosrae, despite the major differences in population genetics and environment.

Genome-wide RNA expression data provide a detailed view of an organism's biological state; hence, a dataset measuring expression variation between genetically diverse individuals (eQTL data) may provide important insights into the genetics of complex traits. However, with data from a relatively small number of individuals, it is difficult to distinguish true causal polymorphisms from the large number of possibilities. The problem is particularly challenging in populations with significant linkage disequilibrium, where traits are often linked to large chromosomal regions containing many genes. Here, we present a novel method, Lirnet, that automatically learns a regulatory potential for each sequence polymorphism, estimating how likely it is to have a significant effect on gene expression. This regulatory potential is defined in terms of "regulatory features"-including the function of the gene and the conservation, type, and position of genetic polymorphisms-that are available for any organism. The extent to which the different features influence the regulatory potential is learned automatically, making Lirnet readily applicable to different datasets, organisms, and feature sets. We apply Lirnet both to the human HapMap eQTL dataset and to a yeast eQTL dataset and provide statistical and biological results demonstrating that Lirnet produces significantly better regulatory programs than other recent approaches. We demonstrate in the yeast data that Lirnet can correctly suggest a specific causal sequence variation within a large, linked chromosomal region. In one example, Lirnet uncovered a novel, experimentally validated connection between Puf3-a sequence-specific RNA binding protein-and P-bodies-cytoplasmic structures that regulate translation and RNA stability-as well as the particular causative polymorphism, a SNP in Mkt1, that induces the variation in the pathway.