Distinguishing between recent balancing selection and incomplete sweep using deep neural networks

Temporal challenges in detecting balancing selection from population genomic data

The role of balancing selection in maintaining genetic variation remains an open question in population genetics. Recent years have seen numerous studies identifying candidate loci potentially experiencing balancing selection, most predominantly in human populations. There are however numerous alternative evolutionary processes that may leave similar patterns of variation, thereby potentially confounding inference, and the expected signatures of balancing selection additionally change in a temporal fashion. Here we use forward-in-time simulations to quantify expected statistical power to detect balancing selection using both site frequency spectrum- and linkage disequilibrium-based methods under a variety of evolutionarily realistic null models. We find that whilst site frequency spectrum-based methods have little power immediately after a balanced mutation begins segregating, power increases with time since the introduction of the balanced allele. Conversely, linkage disequilibrium-based methods have considerable power whilst the allele is young, and power dissipates rapidly as the time since introduction increases. Taken together, this suggests that site frequency spectrum-based methods are most effective at detecting long-term balancing selection (>25N generations since the introduction of the balanced allele) whilst linkage disequilibrium-based methods are effective over much shorter timescales (<1N generations), thereby leaving a large time frame over which current methods have little power to detect the action of balancing selection. Finally, we investigate the extent to which alternative evolutionary processes may mimic these patterns, and demonstrate the need for caution in attempting to distinguish the signatures of balancing selection from those of both neutral processes (e.g. population structure and admixture) as well as of alternative selective processes (e.g. partial selective sweeps).

The Impact of Patterns in Linkage Disequilibrium and Sequencing Quality on the Imprint of Balancing Selection

Regions under balancing selection are characterized by dense polymorphisms and multiple persistent haplotypes, along with other sequence complexities. Successful identification of these patterns depends on both the statistical approach and the quality of sequencing. To address this challenge, at first, a new statistical method called LD-ABF was developed, employing efficient Bayesian techniques to effectively test for balancing selection. LD-ABF demonstrated the most robust detection of selection in a variety of simulation scenarios, compared against a range of existing tests/tools (Tajima's D, HKA, Dng, BetaScan, and BalLerMix). Furthermore, the impact of the quality of sequencing on detection of balancing selection was explored, as well, using: (i) SNP genotyping and exome data, (ii) targeted high-resolution HLA genotyping (IHIW), and (iii) whole-genome long-read sequencing data (Pangenome). In the analysis of SNP genotyping and exome data, we identified known targets and 38 new selection signatures in genes not previously linked to balancing selection. To further investigate the impact of sequencing quality on detection of balancing selection, a detailed investigation of the MHC was performed with high-resolution HLA typing data. Higher quality sequencing revealed the HLA-DQ genes consistently demonstrated strong selection signatures otherwise not observed from the sparser SNP array and exome data. The HLA-DQ selection signature was also replicated in the Pangenome samples using considerably less samples but, with high-quality long-read sequence data. The improved statistical method, coupled with higher quality sequencing, leads to more consistent identification of selection and enhanced localization of variants under selection, particularly in complex regions.

Harnessing deep learning for population genetic inference

In population genetics, the emergence of large-scale genomic data for various species and populations has provided new opportunities to understand the evolutionary forces that drive genetic diversity using statistical inference. However, the era of population genomics presents new challenges in analysing the massive amounts of genomes and variants. Deep learning has demonstrated state-of-the-art performance for numerous applications involving large-scale data. Recently, deep learning approaches have gained popularity in population genetics; facilitated by the advent of massive genomic data sets, powerful computational hardware and complex deep learning architectures, they have been used to identify population structure, infer demographic history and investigate natural selection. Here, we introduce common deep learning architectures and provide comprehensive guidelines for implementing deep learning models for population genetic inference. We also discuss current challenges and future directions for applying deep learning in population genetics, focusing on efficiency, robustness and interpretability.

Bw4 ligand and direct T-cell receptor binding induced selection on HLA A and B alleles

Introduction

The HLA region is the hallmark of balancing selection, argued to be driven by the pressure to present a wide variety of viral epitopes. As such selection on the peptide-binding positions has been proposed to drive HLA population genetics. MHC molecules also directly binds to the T-Cell Receptor and killer cell immunoglobulin-like receptors (KIR).

Methods

We here combine the HLA allele frequencies in over six-million Hematopoietic Stem Cells (HSC) donors with a novel machine-learning-based method to predict allele frequency.

Results

We show for the first time that allele frequency can be predicted from their sequences. This prediction yields a natural measure for selection. The strongest selection is affecting KIR binding regions, followed by the peptide-binding cleft. The selection from the direct interaction with the KIR and TCR is centered on positively charged residues (mainly Arginine), and some positions in the peptide-binding cleft are not associated with the allele frequency, especially Tyrosine residues.

Discussion

These results suggest that the balancing selection for peptide presentation is combined with a positive selection for KIR and TCR binding.

On convolutional neural networks for selection inference: Revealing the effect of preprocessing on model learning and the capacity to discover novel patterns

A central challenge in population genetics is the detection of genomic footprints of selection. As machine learning tools including convolutional neural networks (CNNs) have become more sophisticated and applied more broadly, these provide a logical next step for increasing our power to learn and detect such patterns; indeed, CNNs trained on simulated genome sequences have recently been shown to be highly effective at this task. Unlike previous approaches, which rely upon human-crafted summary statistics, these methods are able to be applied directly to raw genomic data, allowing them to potentially learn new signatures that, if well-understood, could improve the current theory surrounding selective sweeps. Towards this end, we examine a representative CNN from the literature, paring it down to the minimal complexity needed to maintain comparable performance; this low-complexity CNN allows us to directly interpret the learned evolutionary signatures. We then validate these patterns in more complex models using metrics that evaluate feature importance. Our findings reveal that preprocessing steps, which determine how the population genetic data is presented to the model, play a central role in the learned prediction method. This results in models that mimic previously-defined summary statistics; in one case, the summary statistic itself achieves similarly high accuracy. For evolutionary processes that are less well understood than selective sweeps, we hope this provides an initial framework for using CNNs in ways that go beyond simply achieving high classification performance. Instead, we propose that CNNs might be useful as tools for learning novel patterns that can translate to easy-to-implement summary statistics available to a wider community of researchers.

Tensor Decomposition-based Feature Extraction and Classification to Detect Natural Selection from Genomic Data

Inferences of adaptive events are important for learning about traits, such as human digestion of lactose after infancy and the rapid spread of viral variants. Early efforts toward identifying footprints of natural selection from genomic data involved development of summary statistic and likelihood methods. However, such techniques are grounded in simple patterns or theoretical models that limit the complexity of settings they can explore. Due to the renaissance in artificial intelligence, machine learning methods have taken center stage in recent efforts to detect natural selection, with strategies such as convolutional neural networks applied to images of haplotypes. Yet, limitations of such techniques include estimation of large numbers of model parameters under nonconvex settings and feature identification without regard to location within an image. An alternative approach is to use tensor decomposition to extract features from multidimensional data although preserving the latent structure of the data, and to feed these features to machine learning models. Here, we adopt this framework and present a novel approach termed T-REx, which extracts features from images of haplotypes across sampled individuals using tensor decomposition, and then makes predictions from these features using classical machine learning methods. As a proof of concept, we explore the performance of T-REx on simulated neutral and selective sweep scenarios and find that it has high power and accuracy to discriminate sweeps from neutrality, robustness to common technical hurdles, and easy visualization of feature importance. Therefore, T-REx is a powerful addition to the toolkit for detecting adaptive processes from genomic data.

Genetic and clinical features of familial mediterranean fever (FMF) in a homogeneous cohort of patients from South-Eastern Italy

Familial Mediterranean Fever (FMF) is linked with the MEFV gene and is the commonest among monogenic autoinflammatory diseases, with high prevalence in the Mediterranean basin. Although the clinical presentation of FMF has a major role in diagnosis, genotype/phenotype correlations and the role of "benign" gene variants (as R202Q) appear highly variable and incompletely clear, making difficult to select the most effective strategy in the management of patients. Aim of the present study was to investigate the clinical presentation and the genetic background in a homogenous cohort of patients from Apulia (south eastern Italy). We investigated 217 patients with a clinical suspect of autoinflammatory diseases, who were characterized for the occurrence of specific symptoms and with next generation sequencing by a 4-gene panel including MEFV, MVK, NLRP3 and TNFRSF1A. A genetic change was identified in 122 (53.7%) patients, with 161 different MEFV variants recorded in 100 individuals, 10 variants in NLRP3, and 6 each in TNFRSF1A and MVK. The benign variant R202Q was largely prevalent (41.6% of all MEFV variants). When patients were selected according the number of pathogenic MEFV variants (0, 1, or 2 pathogenic variants), results failed to show significant links between the frequency of symptoms and the number of pathogenic variants. Only family history and Pras score (indicative for severity of disease) predicted the presence of pathogenic variants, as compared with carriers of variants considered of uncertain significance or benign. Fever >38 °C and arthralgias appeared more frequently in R202Q-positive patients than in non-R202Q carriers. These two subgroups showed comparable duration of fever, occurrence of myalgia, abdominal and chest pain, Pras, and IFFS scores. In conclusion, results confirm that FMF manifests in mild form in non-middle eastern patients. This possibility partly affects the reliability of clinical criteria/scores. Furthermore, the presence of the R202Q variant might not be completely neutral in selected groups of patients.

INTERPRETING GENERATIVE ADVERSARIAL NETWORKS TO INFER NATURAL SELECTION FROM GENETIC DATA

Understanding natural selection in humans and other species is a major focus for the use of machine learning in population genetics. Existing methods rely on computationally intensive simulated training data. Unlike efficient neutral coalescent simulations for demographic inference, realistic simulations of selection typically requires slow forward simulations. Because there are many possible modes of selection, a high dimensional parameter space must be explored, with no guarantee that the simulated models are close to the real processes. Mismatches between simulated training data and real test data can lead to incorrect inference. Finally, it is difficult to interpret trained neural networks, leading to a lack of understanding about what features contribute to classification. Here we develop a new approach to detect selection that requires relatively few selection simulations during training. We use a Generative Adversarial Network (GAN) trained to simulate realistic neutral data. The resulting GAN consists of a generator (fitted demographic model) and a discriminator (convolutional neural network). For a genomic region, the discriminator predicts whether it is "real" or "fake" in the sense that it could have been simulated by the generator. As the "real" training data includes regions that experienced selection and the generator cannot produce such regions, regions with a high probability of being real are likely to have experienced selection. To further incentivize this behavior, we "fine-tune" the discriminator with a small number of selection simulations. We show that this approach has high power to detect selection in simulations, and that it finds regions under selection identified by state-of-the art population genetic methods in three human populations. Finally, we show how to interpret the trained networks by clustering hidden units of the discriminator based on their correlation patterns with known summary statistics. In summary, our approach is a novel, efficient, and powerful way to use machine learning to detect natural selection.

This population does not exist: learning the distribution of evolutionary histories with generative adversarial networks

Numerous studies over the last decade have demonstrated the utility of machine learning methods when applied to population genetic tasks. More recent studies show the potential of deep-learning methods in particular, which allow researchers to approach problems without making prior assumptions about how the data should be summarized or manipulated, instead learning their own internal representation of the data in an attempt to maximize inferential accuracy. One type of deep neural network, called Generative Adversarial Networks (GANs), can even be used to generate new data, and this approach has been used to create individual artificial human genomes free from privacy concerns. In this study, we further explore the application of GANs in population genetics by designing and training a network to learn the statistical distribution of population genetic alignments (i.e. data sets consisting of sequences from an entire population sample) under several diverse evolutionary histories-the first GAN capable of performing this task. After testing multiple different neural network architectures, we report the results of a fully differentiable Deep-Convolutional Wasserstein GAN with gradient penalty that is capable of generating artificial examples of population genetic alignments that successfully mimic key aspects of the training data, including the site-frequency spectrum, differentiation between populations, and patterns of linkage disequilibrium. We demonstrate consistent training success across various evolutionary models, including models of panmictic and subdivided populations, populations at equilibrium and experiencing changes in size, and populations experiencing either no selection or positive selection of various strengths, all without the need for extensive hyperparameter tuning. Overall, our findings highlight the ability of GANs to learn and mimic population genetic data and suggest future areas where this work can be applied in population genetics research that we discuss herein.

Localizing Post-Admixture Adaptive Variants with Object Detection on Ancestry-Painted Chromosomes

Gene flow between previously differentiated populations during the founding of an admixed or hybrid population has the potential to introduce adaptive alleles into the new population. If the adaptive allele is common in one source population, but not the other, then as the adaptive allele rises in frequency in the admixed population, genetic ancestry from the source containing the adaptive allele will increase nearby as well. Patterns of genetic ancestry have therefore been used to identify post-admixture positive selection in humans and other animals, including examples in immunity, metabolism, and animal coloration. A common method identifies regions of the genome that have local ancestry "outliers" compared with the distribution across the rest of the genome, considering each locus independently. However, we lack theoretical models for expected distributions of ancestry under various demographic scenarios, resulting in potential false positives and false negatives. Further, ancestry patterns between distant sites are often not independent. As a result, current methods tend to infer wide genomic regions containing many genes as under selection, limiting biological interpretation. Instead, we develop a deep learning object detection method applied to images generated from local ancestry-painted genomes. This approach preserves information from the surrounding genomic context and avoids potential pitfalls of user-defined summary statistics. We find the method is robust to a variety of demographic misspecifications using simulated data. Applied to human genotype data from Cabo Verde, we localize a known adaptive locus to a single narrow region compared with multiple or long windows obtained using two other ancestry-based methods.

Deep Learning in Population Genetics

Population genetics is transitioning into a data-driven discipline thanks to the availability of large-scale genomic data and the need to study increasingly complex evolutionary scenarios. With likelihood and Bayesian approaches becoming either intractable or computationally unfeasible, machine learning, and in particular deep learning, algorithms are emerging as popular techniques for population genetic inferences. These approaches rely on algorithms that learn non-linear relationships between the input data and the model parameters being estimated through representation learning from training data sets. Deep learning algorithms currently employed in the field comprise discriminative and generative models with fully connected, convolutional, or recurrent layers. Additionally, a wide range of powerful simulators to generate training data under complex scenarios are now available. The application of deep learning to empirical data sets mostly replicates previous findings of demography reconstruction and signals of natural selection in model organisms. To showcase the feasibility of deep learning to tackle new challenges, we designed a branched architecture to detect signals of recent balancing selection from temporal haplotypic data, which exhibited good predictive performance on simulated data. Investigations on the interpretability of neural networks, their robustness to uncertain training data, and creative representation of population genetic data, will provide further opportunities for technological advancements in the field.

dnadna: a deep learning framework for population genetics inference

MOTIVATION

We present dnadna, a flexible python-based software for deep learning inference in population genetics. It is task-agnostic and aims at facilitating the development, reproducibility, dissemination and re-usability of neural networks designed for population genetic data.

RESULTS

dnadna defines multiple user-friendly workflows. First, users can implement new architectures and tasks, while benefiting from dnadna utility functions, training procedure and test environment, which saves time and decreases the likelihood of bugs. Second, the implemented networks can be re-optimized based on user-specified training sets and/or tasks. Newly implemented architectures and pre-trained networks are easily shareable with the community for further benchmarking or other applications. Finally, users can apply pre-trained networks in order to predict evolutionary history from alternative real or simulated genetic datasets, without requiring extensive knowledge in deep learning or coding in general. dnadna comes with a peer-reviewed, exchangeable neural network, allowing demographic inference from SNP data, that can be used directly or retrained to solve other tasks. Toy networks are also available to ease the exploration of the software, and we expect that the range of available architectures will keep expanding thanks to community contributions.

AVAILABILITY AND IMPLEMENTATION

dnadna is a Python (≥3.7) package, its repository is available at gitlab.com/mlgenetics/dnadna and its associated documentation at mlgenetics.gitlab.io/dnadna/.

Deciphering signatures of natural selection via deep learning

Identifying genomic regions influenced by natural selection provides fundamental insights into the genetic basis of local adaptation. However, it remains challenging to detect loci under complex spatially varying selection. We propose a deep learning-based framework, DeepGenomeScan, which can detect signatures of spatially varying selection. We demonstrate that DeepGenomeScan outperformed principal component analysis- and redundancy analysis-based genome scans in identifying loci underlying quantitative traits subject to complex spatial patterns of selection. Noticeably, DeepGenomeScan increases statistical power by up to 47.25% under nonlinear environmental selection patterns. We applied DeepGenomeScan to a European human genetic dataset and identified some well-known genes under selection and a substantial number of clinically important genes that were not identified by SPA, iHS, Fst and Bayenv when applied to the same dataset.

Evolution in small steps and giant leaps

The first Editor of Evolution was Ernst Mayr. His foreword to the first issue of Evolution published in 1947 framed evolution as a "problem of interaction" that was just beginning to be studied in this broad context. First, I explore progress and prospects on understanding the subsidiary interactions identified by Mayr, including interactions between parts of organisms, between individuals and populations, between species, and between the organism and its abiotic environment. Mayr's overall "problem of interaction" framework is examined in the context of coevolution within and among levels of biological organization. This leads to a comparison in the relative roles of biotic versus abiotic agents of selection and fluctuating versus directional selection, followed by stabilizing selection in shaping the genomic architecture of adaptation. Oligogenic architectures may be typical for traits shaped more by fluctuating selection and biotic selection. Conversely, polygenic architectures may be typical for traits shaped more by directional followed by stabilizing selection and abiotic selection. The distribution of effect sizes and turnover dynamics of adaptive alleles in these scenarios deserves further study. Second, I review two case studies on the evolution of acquired toxicity in animals, one involving cardiac glycosides obtained from plants and one involving bacterial virulence factors horizontally transferred to animals. The approaches used in these studies and the results gained directly flow from Mayr's vision of an evolutionary biology that revolves around the "problem of interaction."

A Study of the Recent Trends of Immunology: Key Challenges, Domains, Applications, Datasets, and Future Directions

The human immune system is very complex. Understanding it traditionally required specialized knowledge and expertise along with years of study. However, in recent times, the introduction of technologies such as AIoMT (Artificial Intelligence of Medical Things), genetic intelligence algorithms, smart immunological methodologies, etc., has made this process easier. These technologies can observe relations and patterns that humans do and recognize patterns that are unobservable by humans. Furthermore, these technologies have also enabled us to understand better the different types of cells in the immune system, their structures, their importance, and their impact on our immunity, particularly in the case of debilitating diseases such as cancer. The undertaken study explores the AI methodologies currently in the field of immunology. The initial part of this study explains the integration of AI in healthcare and how it has changed the face of the medical industry. It also details the current applications of AI in the different healthcare domains and the key challenges faced when trying to integrate AI with healthcare, along with the recent developments and contributions in this field by other researchers. The core part of this study is focused on exploring the most common classifications of health diseases, immunology, and its key subdomains. The later part of the study presents a statistical analysis of the contributions in AI in the different domains of immunology and an in-depth review of the machine learning and deep learning methodologies and algorithms that can and have been applied in the field of immunology. We have also analyzed a list of machine learning and deep learning datasets about the different subdomains of immunology. Finally, in the end, the presented study discusses the future research directions in the field of AI in immunology and provides some possible solutions for the same.

BalLeRMix +: mixture model approaches for robust joint identification of both positive selection and long-term balancing selection

SUMMARY

The growing availability of genomewide polymorphism data has fueled interest in detecting diverse selective processes affecting population diversity. However, no model-based approaches exist to jointly detect and distinguish the two complementary processes of balancing and positive selection. We extend the BalLeRMix  B-statistic framework described in Cheng and DeGiorgio (2020) for detecting balancing selection and present BalLeRMix+, which implements five B statistic extensions based on mixture models to robustly identify both types of selection. BalLeRMix+ is implemented in Python and computes the composite likelihood ratios and associated model parameters for each genomic test position.

AVAILABILITY AND IMPLEMENTATION

BalLeRMix+ is freely available at https://github.com/bioXiaoheng/BallerMixPlus.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Detecting adaptive introgression in human evolution using convolutional neural networks

Studies in a variety of species have shown evidence for positively selected variants introduced into a population via introgression from another, distantly related population—a process known as adaptive introgression. However, there are few explicit frameworks for jointly modelling introgression and positive selection, in order to detect these variants using genomic sequence data. Here, we develop an approach based on convolutional neural networks (CNNs). CNNs do not require the specification of an analytical model of allele frequency dynamics and have outperformed alternative methods for classification and parameter estimation tasks in various areas of population genetics. Thus, they are potentially well suited to the identification of adaptive introgression. Using simulations, we trained CNNs on genotype matrices derived from genomes sampled from the donor population, the recipient population and a related non-introgressed population, in order to distinguish regions of the genome evolving under adaptive introgression from those evolving neutrally or experiencing selective sweeps. Our CNN architecture exhibits 95% accuracy on simulated data, even when the genomes are unphased, and accuracy decreases only moderately in the presence of heterosis. As a proof of concept, we applied our trained CNNs to human genomic datasets—both phased and unphased—to detect candidates for adaptive introgression that shaped our evolutionary history.

Understanding the Adaptive Evolutionary Histories of South American Ancient and Present-Day Populations via Genomics

The South American continent is remarkably diverse in its ecological zones, spanning the Amazon rainforest, the high-altitude Andes, and Tierra del Fuego. Yet the original human populations of the continent successfully inhabited all these zones, well before the buffering effects of modern technology. Therefore, it is likely that the various cultures were successful, in part, due to positive natural selection that allowed them to successfully establish populations for thousands of years. Detecting positive selection in these populations is still in its infancy, as the ongoing effects of European contact have decimated many of these populations and introduced gene flow from outside of the continent. In this review, we explore hypotheses of possible human biological adaptation, methods to identify positive selection, the utilization of ancient DNA, and the integration of modern genomes through the identification of genomic tracts that reflect the ancestry of the first populations of the Americas.