Inferring genome-wide recombination landscapes from advanced intercross lines: application to yeast crosses

Approximate Bayesian computational methods to estimate the strength of divergent selection in population genomics models

Statistical estimation of parameters in large models of evolutionary processes is often too computationally inefficient to pursue using exact model likelihoods, even with single-nucleotide polymorphism (SNP) data, which offers a way to reduce the size of genetic data while retaining relevant information. Approximate Bayesian Computation (ABC) to perform statistical inference about parameters of large models takes the advantage of simulations to bypass direct evaluation of model likelihoods. We develop a mechanistic model to simulate forward-in-time divergent selection with variable migration rates, modes of reproduction (sexual, asexual), length and number of migration-selection cycles. We investigate the computational feasibility of ABC to perform statistical inference and study the quality of estimates on the position of loci under selection and the strength of selection. To expand the parameter space of positions under selection, we enhance the model by implementing an outlier scan on summarized observed data. We evaluate the usefulness of summary statistics well-known to capture the strength of selection, and assess their informativeness under divergent selection. We also evaluate the effect of genetic drift with respect to an idealized deterministic model with single-locus selection. We discuss the role of the recombination rate as a confounding factor in estimating the strength of divergent selection, and emphasize its importance in break down of linkage disequilibrium (LD). We answer the question for which part of the parameter space of the model we recover strong signal for estimating the selection, and determine whether population differentiation-based summary statistics or LD-based summary statistics perform well in estimating selection.

Adaptation in Outbred Sexual Yeast is Repeatable, Polygenic and Favors Rare Haplotypes

We carried out a 200 generation Evolve and Resequence (E&R) experiment initiated from an outbred diploid recombined 18-way synthetic base population. Replicate populations were evolved at large effective population sizes (>105 individuals), exposed to several different chemical challenges over 12 weeks of evolution, and whole-genome resequenced. Weekly forced outcrossing resulted in an average between adjacent-gene per cell division recombination rate of ∼0.0008. Despite attempts to force weekly sex, roughly half of our populations evolved cheaters and appear to be evolving asexually. Focusing on seven chemical stressors and 55 total evolved populations that remained sexual we observed large fitness gains and highly repeatable patterns of genome-wide haplotype change within chemical challenges, with limited levels of repeatability across chemical treatments. Adaptation appears highly polygenic with almost the entire genome showing significant and consistent patterns of haplotype change with little evidence for long-range linkage disequilibrium in a subset of populations for which we sequenced haploid clones. That is, almost the entire genome is under selection or drafting with selected sites. At any given locus adaptation was almost always dominated by one of the 18 founder's alleles, with that allele varying spatially and between treatments, suggesting that selection acts primarily on rare variants private to a founder or haplotype blocks harboring multiple mutations.

A widespread inversion polymorphism conserved among Saccharomyces species is caused by recurrent homogenization of a sporulation gene family

Saccharomyces genomes are highly collinear and show relatively little structural variation, both within and between species of this yeast genus. We investigated the only common inversion polymorphism known in S. cerevisiae, which affects a 24-kb 'flip/flop' region containing 15 genes near the centromere of chromosome XIV. The region exists in two orientations, called reference (REF) and inverted (INV). Meiotic recombination in this region is suppressed in crosses between REF and INV orientation strains such as the BY x RM cross. We find that the inversion polymorphism is at least 17 million years old because it is conserved across the genus Saccharomyces. However, the REF and INV isomers are not ancient alleles but are continually being re-created by re-inversion of the region within each species. Inversion occurs due to continual homogenization of two almost identical 4-kb sequences that form an inverted repeat (IR) at the ends of the flip/flop region. The IR consists of two pairs of genes that are specifically and strongly expressed during the late stages of sporulation. We show that one of these gene pairs, YNL018C/YNL034W, codes for a protein that is essential for spore formation. YNL018C and YNL034W are the founder members of a gene family, Centroid, whose members in other Saccharomycetaceae species evolve fast, duplicate frequently, and are preferentially located close to centromeres. We tested the hypothesis that Centroid genes are a meiotic drive system, but found no support for this idea.

Estimating the time since admixture from phased and unphased molecular data

After admixture, recombination breaks down genomic blocks of contiguous ancestry. The breakdown of these blocks forms a new “molecular clock” that ticks at a much faster rate than the mutation clock, enabling accurate dating of admixture events in the recent past. However, existing theory on the breakdown of these blocks, or the accumulation of delineations between blocks, so‐called “junctions”, has mostly been limited to using regularly spaced markers on phased data. Here, we present an extension to the theory of junctions using the ancestral recombination graph that describes the expected number of junctions for any distribution of markers along the genome. Furthermore, we provide a new framework to infer the time since admixture using unphased data. We demonstrate both the phased and unphased methods on simulated data and show that our new extensions have improved accuracy with respect to previous methods, especially for smaller population sizes and more ancient admixture times. Lastly, we demonstrate the applicability of our method on three empirical data sets, including labcrosses of yeast (Saccharomyces cerevisae) and two case studies of hybridization in swordtail fish and Populus trees.

Intragenic repeat expansion in the cell wall protein gene HPF1 controls yeast chronological aging

Aging varies among individuals due to both genetics and environment, but the underlying molecular mechanisms remain largely unknown. Using a highly recombined Saccharomyces cerevisiae population, we found 30 distinct quantitative trait loci (QTLs) that control chronological life span (CLS) in calorie-rich and calorie-restricted environments and under rapamycin exposure. Calorie restriction and rapamycin extended life span in virtually all genotypes but through different genetic variants. We tracked the two major QTLs to the cell wall glycoprotein genes FLO11 and HPF1 We found that massive expansion of intragenic tandem repeats within the N-terminal domain of HPF1 was sufficient to cause pronounced life span shortening. Life span impairment by HPF1 was buffered by rapamycin but not by calorie restriction. The HPF1 repeat expansion shifted yeast cells from a sedentary to a buoyant state, thereby increasing their exposure to surrounding oxygen. The higher oxygenation altered methionine, lipid, and purine metabolism, and inhibited quiescence, which explains the life span shortening. We conclude that fast-evolving intragenic repeat expansions can fundamentally change the relationship between cells and their environment with profound effects on cellular lifestyle and longevity.

Population genomics and evolution of a fungal pathogen after releasing exotic strains to control insect pests for 20 years

Entomopathogenic fungi are one of the key regulators of insect populations in nature. Some species such as Beauveria bassiana with a wide host range have been developed as promising alternatives to chemical insecticides for the biocontrol of insect pests. However, the long-term persistence of the released strains, the effect on non-target hosts and local fungal populations remains elusive, but they are considerable concerns with respect to environmental safety. Here we report the temporal features of the Beauveria population genomics and evolution over 20 years after releasing exotic strains to control pine caterpillar pests. We found that the isolates within the biocontrol site were mostly of clonal origins. The released strains could persist in the environment for a long time but with low recovery rates. Similar to the reoccurrence of host jumping by local isolates, the infection of non-target insects by the released strains was evident to endemically occur in association with host seasonality. No obvious dilution effect on local population structure was evident by the releases. However, the population was largely replaced by genetically divergent isolates once per decade but evolved with a pattern of balancing selection and towards expansion through adaptation, non-random outcrossing and isolate migration. This study not only unveils the real-time features of entomopathogenic fungal population genomics and evolution but also provides added values to alleviate the concerns of environmental safety regarding the biocontrol application of mycoinsecticides.

Experimental Evolution of Yeast for High-Temperature Tolerance

Thermotolerance is a polygenic trait that contributes to cell survival and growth under unusually high temperatures. Although some genes associated with high-temperature growth (Htg+) have been identified, how cells accumulate mutations to achieve prolonged thermotolerance is still mysterious. Here, we conducted experimental evolution of a Saccharomyces cerevisiae laboratory strain with stepwise temperature increases for it to grow at 42 °C. Whole genome resequencing of 14 evolved strains and the parental strain revealed a total of 153 mutations in the evolved strains, including single nucleotide variants, small INDELs, and segmental duplication/deletion events. Some mutations persisted from an intermediate temperature to 42 °C, so they might be Htg+ mutations. Functional categorization of mutations revealed enrichment of exonic mutations in the SWI/SNF complex and F-type ATPase, pointing to their involvement in high-temperature tolerance. In addition, multiple mutations were found in a general stress-associated signal transduction network consisting of Hog1 mediated pathway, RAS-cAMP pathway, and Rho1-Pkc1 mediated cell wall integrity pathway, implying that cells can achieve Htg+ partly through modifying existing stress regulatory mechanisms. Using pooled segregant analysis of five Htg+ phenotype-orientated pools, we inferred causative mutations for growth at 42 °C and identified those mutations with stronger impacts on the phenotype. Finally, we experimentally validated a number of the candidate Htg+ mutations. This study increased our understanding of the genetic basis of yeast tolerance to high temperature.

Environment-dependent pleiotropic effects of mutations on the maximum growth rate r and carrying capacity K of population growth

Maximum growth rate per individual (r) and carrying capacity (K) are key life-history traits that together characterize the density-dependent population growth and therefore are crucial parameters of many ecological and evolutionary theories such as r/K selection. Although r and K are generally thought to correlate inversely, both r/K tradeoffs and trade-ups have been observed. Nonetheless, neither the conditions under which each of these relationships occur nor the causes of these relationships are fully understood. Here, we address these questions using yeast as a model system. We estimated r and K using the growth curves of over 7,000 yeast recombinants in nine environments and found that the r-K correlation among genotypes changes from 0.53 to -0.52 with the rise of environment quality, measured by the mean r of all genotypes in the environment. We respectively mapped quantitative trait loci (QTLs) for r and K in each environment. Many QTLs simultaneously influence r and K, but the directions of their effects are environment dependent such that QTLs tend to show concordant effects on the two traits in poor environments but antagonistic effects in rich environments. We propose that these contrasting trends are generated by the relative impacts of two factors-the tradeoff between the speed and efficiency of ATP production and the energetic cost of cell maintenance relative to reproduction-and demonstrate an agreement between model predictions and empirical observations. These results reveal and explain the complex environment dependency of the r-K relationship, which bears on many ecological and evolutionary phenomena and has biomedical implications.

Role of Cis, Trans, and Inbreeding Effects on Meiotic Recombination in Saccharomyces cerevisiae

Meiotic recombination is a major driver of genome evolution by creating new genetic combinations. To probe the factors driving variability of meiotic recombination, we used a high-throughput method to measure recombination rates in hybrids between SK1 and a total of 26 Saccharomyces cerevisiae strains from different geographic origins and habitats. Fourteen intervals were monitored for each strain, covering chromosomes VI and XI entirely, and part of chromosome I. We found an average number of crossovers per chromosome ranging between 1.0 and 9.5 across strains ("domesticated" or not), which is higher than the average between 0.5 and 1.5 found in most organisms. In the different intervals analyzed, recombination showed up to ninefold variation across strains but global recombination landscapes along chromosomes varied less. We also built an incomplete diallel experiment to measure recombination rates in one region of chromosome XI in 10 different crosses involving five parental strains. Our overall results indicate that recombination rate is increasingly positively correlated with sequence similarity between homologs (i) in DNA double-strand-break-rich regions within intervals, (ii) in entire intervals, and (iii) at the whole genome scale. Therefore, these correlations cannot be explained by cis effects only. We also estimated that cis and trans effects explained 38 and 17%, respectively, of the variance of recombination rate. In addition, by using a quantitative genetics analysis, we identified an inbreeding effect that reduces recombination rate in homozygous genotypes, while other interaction effects (specific combining ability) or additive effects (general combining ability) are found to be weak. Finally, we measured significant crossover interference in some strains, and interference intensity was positively correlated with crossover number.

Identifying Drivers of Parallel Evolution: A Regression Model Approach

Parallel evolution, defined as identical changes arising in independent populations, is often attributed to similar selective pressures favoring the fixation of identical genetic changes. However, some level of parallel evolution is also expected if mutation rates are heterogeneous across regions of the genome. Theory suggests that mutation and selection can have equal impacts on patterns of parallel evolution; however, empirical studies have yet to jointly quantify the importance of these two processes. Here, we introduce several statistical models to examine the contributions of mutation and selection heterogeneity to shaping parallel evolutionary changes at the gene-level. Using this framework, we analyze published data from forty experimentally evolved Saccharomyces cerevisiae populations. We can partition the effects of a number of genomic variables into those affecting patterns of parallel evolution via effects on the rate of arising mutations, and those affecting the retention versus loss of the arising mutations (i.e., selection). Our results suggest that gene-to-gene heterogeneity in both mutation and selection, associated with gene length, recombination rate, and number of protein domains drive parallel evolution at both synonymous and nonsynonymous sites. While there are still a number of parallel changes that are not well described, we show that allowing for heterogeneous rates of mutation and selection can provide improved predictions of the prevalence and degree of parallel evolution.

Mapping Quantitative Trait Loci in Yeast

Natural Saccharomyces strains isolated from the wild differ quantitatively in molecular and organismal phenotypes. Quantitative trait loci (QTL) mapping is a powerful approach for identifying sequence variants that alter gene function. In yeast, QTL mapping has been used in designed crosses to map functional polymorphisms. This approach, outlined here, is often the first step in understanding the molecular basis of quantitative traits. New large-scale sequencing surveys have the potential to directly associate genotypes with organismal phenotypes, providing a broader catalog of causative genetic variants. Additional analysis of intermediate phenotypes (e.g., RNA, protein, or metabolite levels) can produce a multilayered and integrated view of individual variation, producing a high-resolution view of the genotype-phenotype map.

Resolving the Complex Genetic Basis of Phenotypic Variation and Variability of Cellular Growth

In all organisms, the majority of traits vary continuously between individuals. Explaining the genetic basis of quantitative trait variation requires comprehensively accounting for genetic and nongenetic factors as well as their interactions. The growth of microbial cells can be characterized by a lag duration, an exponential growth phase, and a stationary phase. Parameters that characterize these growth phases can vary among genotypes (phenotypic variation), environmental conditions (phenotypic plasticity), and among isogenic cells in a given environment (phenotypic variability). We used a high-throughput microscopy assay to map genetic loci determining variation in lag duration and exponential growth rate in growth rate-limiting and nonlimiting glucose concentrations, using segregants from a cross of two natural isolates of the budding yeast, Saccharomyces cerevisiae We find that some quantitative trait loci (QTL) are common between traits and environments whereas some are unique, exhibiting gene-by-environment interactions. Furthermore, whereas variation in the central tendency of growth rate or lag duration is explained by many additive loci, differences in phenotypic variability are primarily the result of genetic interactions. We used bulk segregant mapping to increase QTL resolution by performing whole-genome sequencing of complex mixtures of an advanced intercross mapping population grown in selective conditions using glucose-limited chemostats. We find that sequence variation in the high-affinity glucose transporter HXT7 contributes to variation in growth rate and lag duration. Allele replacements of the entire locus, as well as of a single polymorphic amino acid, reveal that the effect of variation in HXT7 depends on genetic, and allelic, background. Amplifications of HXT7 are frequently selected in experimental evolution in glucose-limited environments, but we find that HXT7 amplifications result in antagonistic pleiotropy that is absent in naturally occurring variants of HXT7 Our study highlights the complex nature of the genotype-to-phenotype map within and between environments.

Contrasting evolutionary genome dynamics between domesticated and wild yeasts

Structural rearrangements have long been recognized as an important source of genetic variation, with implications in phenotypic diversity and disease, yet their detailed evolutionary dynamics remain elusive. Here we use long-read sequencing to generate end-to-end genome assemblies for 12 strains representing major subpopulations of the partially domesticated yeast Saccharomyces cerevisiae and its wild relative Saccharomyces paradoxus. These population-level high-quality genomes with comprehensive annotation enable precise definition of chromosomal boundaries between cores and subtelomeres and a high-resolution view of evolutionary genome dynamics. In chromosomal cores, S. paradoxus shows faster accumulation of balanced rearrangements (inversions, reciprocal translocations and transpositions), whereas S. cerevisiae accumulates unbalanced rearrangements (novel insertions, deletions and duplications) more rapidly. In subtelomeres, both species show extensive interchromosomal reshuffling, with a higher tempo in S. cerevisiae. Such striking contrasts between wild and domesticated yeasts are likely to reflect the influence of human activities on structural genome evolution.

Powerful decomposition of complex traits in a diploid model

Explaining trait differences between individuals is a core and challenging aim of life sciences. Here, we introduce a powerful framework for complete decomposition of trait variation into its underlying genetic causes in diploid model organisms. We sequence and systematically pair the recombinant gametes of two intercrossed natural genomes into an array of diploid hybrids with fully assembled and phased genomes, termed Phased Outbred Lines (POLs). We demonstrate the capacity of this approach by partitioning fitness traits of 6,642 Saccharomyces cerevisiae POLs across many environments, achieving near complete trait heritability and precisely estimating additive (73%), dominance (10%), second (7%) and third (1.7%) order epistasis components. We map quantitative trait loci (QTLs) and find nonadditive QTLs to outnumber (3:1) additive loci, dominant contributions to heterosis to outnumber overdominant, and extensive pleiotropy. The POL framework offers the most complete decomposition of diploid traits to date and can be adapted to most model organisms.

Predicting quantitative traits from genome and phenome with near perfect accuracy

In spite of decades of linkage and association studies and its potential impact on human health, reliable prediction of an individual's risk for heritable disease remains difficult. Large numbers of mapped loci do not explain substantial fractions of heritable variation, leaving an open question of whether accurate complex trait predictions can be achieved in practice. Here, we use a genome sequenced population of ∼7,000 yeast strains of high but varying relatedness, and predict growth traits from family information, effects of segregating genetic variants and growth in other environments with an average coefficient of determination R(2) of 0.91. This accuracy exceeds narrow-sense heritability, approaches limits imposed by measurement repeatability and is higher than achieved with a single assay in the laboratory. Our results prove that very accurate prediction of complex traits is possible, and suggest that additional data from families rather than reference cohorts may be more useful for this purpose.

RIM15 antagonistic pleiotropy is responsible for differences in fermentation and stress response kinetics in budding yeast

Different natural yeast populations have faced dissimilar selective pressures due to the heterogeneous fermentation substrates available around the world; this increases the genetic and phenotypic diversity in Saccharomyces cerevisiae In this context, we expect prominent differences between isolates when exposed to a particular condition, such as wine or sake musts. To better comprehend the mechanisms underlying niche adaptation between two S. cerevisiae isolates obtained from wine and sake fermentation processes, we evaluated fermentative and fungicide resistance phenotypes and identify the molecular origin of such adaptive variation. Multiple regions were associated with fermentation rate under different nitrogen conditions and fungicide resistance, with a single QTL co-localizing in all traits. Analysis around this region identified RIM15 as the causative locus driving fungicide sensitivity, together with efficient nitrogen utilization and glycerol production in the wine strain. A null RIM15 variant confers a greater fermentation rate through the utilization of available glucose instead of its storage. However, this variant has a detrimental effect on fungicide resistance since complex sugars are not synthesized and transported into the membrane. Together, our results reveal the antagonist pleiotropic nature of a RIM15 null variant, positively affecting a series of fermentation related phenotypes, but apparently detrimental in the wild.

Genomic correlates of recombination rate and its variability across eight recombination maps in the western honey bee (Apis mellifera L.)

Background

Meiotic recombination has traditionally been explained based on the structural requirement to stabilize homologous chromosome pairs to ensure their proper meiotic segregation. Competing hypotheses seek to explain the emerging findings of significant heterogeneity in recombination rates within and between genomes, but intraspecific comparisons of genome-wide recombination patterns are rare. The honey bee (Apis mellifera) exhibits the highest rate of genomic recombination among multicellular animals with about five cross-over events per chromatid.

Results

Here, we present a comparative analysis of recombination rates across eight genetic linkage maps of the honey bee genome to investigate which genomic sequence features are correlated with recombination rate and with its variation across the eight data sets, ranging in average marker spacing ranging from 1 Mbp to 120 kbp. Overall, we found that GC content explained best the variation in local recombination rate along chromosomes at the analyzed 100 kbp scale. In contrast, variation among the different maps was correlated to the abundance of microsatellites and several specific tri- and tetra-nucleotides.

Conclusions

The combined evidence from eight medium-scale recombination maps of the honey bee genome suggests that recombination rate variation in this highly recombining genome might be due to the DNA configuration instead of distinct sequence motifs. However, more fine-scale analyses are needed. The empirical basis of eight differing genetic maps allowed for robust conclusions about the correlates of the local recombination rates and enabled the study of the relation between DNA features and variability in local recombination rates, which is particularly relevant in the honey bee genome with its exceptionally high recombination rate.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1281-2) contains supplementary material, which is available to authorized users.

Fine-mapping QTLs in advanced intercross lines and other outbred populations

Quantitative genetic studies in model organisms, particularly in mice, have been extremely successful in identifying chromosomal regions that are associated with a wide variety of behavioral and other traits. However, it is now widely understood that identification of the underlying genes will be far more challenging. In the last few years, a variety of populations have been utilized in an effort to more finely map these chromosomal regions with the goal of identifying specific genes. The common property of these newer populations is that linkage disequilibrium spans relatively short distances, which permits fine-scale mapping resolution. This review focuses on advanced intercross lines (AILs) which are the simplest such population. As originally proposed in 1995 by Darvasi and Soller, an AIL is the product of intercrossing two inbred strains beyond the F2 generation. Unlike recombinant inbred strains, AILs are maintained as outbred populations; brother-sister matings are specifically avoided. Each generation of intercrossing beyond the F2 further degrades linkage disequilibrium between adjacent makers, which allows for fine-scale mapping of quantitative trait loci (QTLs). Advances in genotyping technology and techniques for the statistical analysis of AILs have permitted rapid advances in the application of AILs. We review some of the analytical issues and available software, including QTLRel, EMMA, EMMAX, GEMMA, TASSEL, GRAMMAR, WOMBAT, Mendel, and others.

Genome-wide mapping of cellular traits using yeast

Yeast has long enjoyed superiority as a genetic model because of its short generation time and ease of generating alleles for genetic analysis. However, recent developments of guided nucleases for genome editing in higher eukaryotes, and funding pressures for translational findings, force all model organism communities to reaffirm and rearticulate the advantages of their chosen creature. Here I examine the utility of budding yeast for understanding the genetic basis of cellular traits, using natural variation as well as classical genetic perturbations, and its future prospects compared to undertaking the work in human cell lines. Will yeast remain central, or will it join the likes of phage as an early model that is no longer widely used to answer the pressing questions?

A high-definition view of functional genetic variation from natural yeast genomes

The question of how genetic variation in a population influences phenotypic variation and evolution is of major importance in modern biology. Yet much is still unknown about the relative functional importance of different forms of genome variation and how they are shaped by evolutionary processes. Here we address these questions by population level sequencing of 42 strains from the budding yeast Saccharomyces cerevisiae and its closest relative S. paradoxus. We find that genome content variation, in the form of presence or absence as well as copy number of genetic material, is higher within S. cerevisiae than within S. paradoxus, despite genetic distances as measured in single-nucleotide polymorphisms being vastly smaller within the former species. This genome content variation, as well as loss-of-function variation in the form of premature stop codons and frameshifting indels, is heavily enriched in the subtelomeres, strongly reinforcing the relevance of these regions to functional evolution. Genes affected by these likely functional forms of variation are enriched for functions mediating interaction with the external environment (sugar transport and metabolism, flocculation, metal transport, and metabolism). Our results and analyses provide a comprehensive view of genomic diversity in budding yeast and expose surprising and pronounced differences between the variation within S. cerevisiae and that within S. paradoxus. We also believe that the sequence data and de novo assemblies will constitute a useful resource for further evolutionary and population genomics studies.

High-resolution mapping of complex traits with a four-parent advanced intercross yeast population

A large fraction of human complex trait heritability is due to a high number of variants with small marginal effects and their interactions with genotype and environment. Such alleles are more easily studied in model organisms, where environment, genetic makeup, and allele frequencies can be controlled. Here, we examine the effect of natural genetic variation on heritable traits in a very large pool of baker's yeast from a multiparent 12th generation intercross. We selected four representative founder strains to produce the Saccharomyces Genome Resequencing Project (SGRP)-4X mapping population and sequenced 192 segregants to generate an accurate genetic map. Using these individuals, we mapped 25 loci linked to growth traits under heat stress, arsenite, and paraquat, the majority of which were best explained by a diverging phenotype caused by a single allele in one condition. By sequencing pooled DNA from millions of segregants grown under heat stress, we further identified 34 and 39 regions selected in haploid and diploid pools, respectively, with most of the selection against a single allele. While the most parsimonious model for the majority of loci mapped using either approach was the effect of an allele private to one founder, we could validate examples of pleiotropic effects and complex allelic series at a locus. SGRP-4X is a deeply characterized resource that provides a framework for powerful and high-resolution genetic analysis of yeast phenotypes and serves as a test bed for testing avenues to attack human complex traits.