From disease association to risk assessment: an optimistic view from genome-wide association studies on type 1 diabetes

Unravelling the Genetic Landscape of Hemiplegic Migraine: Exploring Innovative Strategies and Emerging Approaches

Migraine is a severe, debilitating neurovascular disorder. Hemiplegic migraine (HM) is a rare and debilitating neurological condition with a strong genetic basis. Sequencing technologies have improved the diagnosis and our understanding of the molecular pathophysiology of HM. Linkage analysis and sequencing studies in HM families have identified pathogenic variants in ion channels and related genes, including CACNA1A, ATP1A2, and SCN1A, that cause HM. However, approximately 75% of HM patients are negative for these mutations, indicating there are other genes involved in disease causation. In this review, we explored our current understanding of the genetics of HM. The evidence presented herein summarises the current knowledge of the genetics of HM, which can be expanded further to explain the remaining heritability of this debilitating condition. Innovative bioinformatics and computational strategies to cover the entire genetic spectrum of HM are also discussed in this review.

Genomic Machine Learning Meta-regression: Insights on Associations of Study Features With Reported Model Performance

Many studies have been conducted with the goal of correctly predicting diagnostic status of a disorder using the combination of genomic data and machine learning. It is often hard to judge which components of a study led to better results and whether better reported results represent a true improvement or an uncorrected bias inflating performance. We extracted information about the methods used and other differentiating features in genomic machine learning models. We used these features in linear regressions predicting model performance. We tested for univariate and multivariate associations as well as interactions between features. Of the models reviewed, 46% used feature selection methods that can lead to data leakage. Across our models, the number of hyperparameter optimizations reported, data leakage due to feature selection, model type, and modeling an autoimmune disorder were significantly associated with an increase in reported model performance. We found a significant, negative interaction between data leakage and training size. Our results suggest that methods susceptible to data leakage are prevalent among genomic machine learning research, resulting in inflated reported performance. Best practice guidelines that promote the avoidance and recognition of data leakage may help the field avoid biased results.

Genomic Innovation in Early Life Cardiovascular Disease Prevention and Treatment

Cardiovascular disease (CVD) is a leading cause of morbidity and mortality globally. Although CVD events do not typically manifest until older adulthood, CVD develops gradually across the life-course, beginning with the elevation of risk factors observed as early as childhood or adolescence and the emergence of subclinical disease that can occur in young adulthood or midlife. Genomic background, which is determined at zygote formation, is among the earliest risk factors for CVD. With major advances in molecular technology, including the emergence of gene-editing techniques, along with deep whole-genome sequencing and high-throughput array-based genotyping, scientists now have the opportunity to not only discover genomic mechanisms underlying CVD but use this knowledge for the life-course prevention and treatment of these conditions. The current review focuses on innovations in the field of genomics and their applications to monogenic and polygenic CVD prevention and treatment. With respect to monogenic CVD, we discuss how the emergence of whole-genome sequencing technology has accelerated the discovery of disease-causing variants, allowing comprehensive screening and early, aggressive CVD mitigation strategies in patients and their families. We further describe advances in gene editing technology, which might soon make possible cures for CVD conditions once thought untreatable. In relation to polygenic CVD, we focus on recent innovations that leverage findings of genome-wide association studies to identify druggable gene targets and develop predictive genomic models of disease, which are already facilitating breakthroughs in the life-course treatment and prevention of CVD. Gaps in current research and future directions of genomics studies are also discussed. In aggregate, we hope to underline the value of leveraging genomics and broader multiomics information for characterizing CVD conditions, work which promises to expand precision approaches for the life-course prevention and treatment of CVD.

Deep Learning Framework for Complex Disease Risk Prediction Using Genomic Variations

Genome-wide association studies have proven their ability to improve human health outcomes by identifying genotypes associated with phenotypes. Various works have attempted to predict the risk of diseases for individuals based on genotype data. This prediction can either be considered as an analysis model that can lead to a better understanding of gene functions that underlie human disease or as a black box in order to be used in decision support systems and in early disease detection. Deep learning techniques have gained more popularity recently. In this work, we propose a deep-learning framework for disease risk prediction. The proposed framework employs a multilayer perceptron (MLP) in order to predict individuals' disease status. The proposed framework was applied to the Wellcome Trust Case-Control Consortium (WTCCC), the UK National Blood Service (NBS) Control Group, and the 1958 British Birth Cohort (58C) datasets. The performance comparison of the proposed framework showed that the proposed approach outperformed the other methods in predicting disease risk, achieving an area under the curve (AUC) up to 0.94.

Learning high-order interactions for polygenic risk prediction

Within the framework of precision medicine, the stratification of individual genetic susceptibility based on inherited DNA variation has paramount relevance. However, one of the most relevant pitfalls of traditional Polygenic Risk Scores (PRS) approaches is their inability to model complex high-order non-linear SNP-SNP interactions and their effect on the phenotype (e.g. epistasis). Indeed, they incur in a computational challenge as the number of possible interactions grows exponentially with the number of SNPs considered, affecting the statistical reliability of the model parameters as well. In this work, we address this issue by proposing a novel PRS approach, called High-order Interactions-aware Polygenic Risk Score (hiPRS), that incorporates high-order interactions in modeling polygenic risk. The latter combines an interaction search routine based on frequent itemsets mining and a novel interaction selection algorithm based on Mutual Information, to construct a simple and interpretable weighted model of user-specified dimensionality that can predict a given binary phenotype. Compared to traditional PRSs methods, hiPRS does not rely on GWAS summary statistics nor any external information. Moreover, hiPRS differs from Machine Learning-based approaches that can include complex interactions in that it provides a readable and interpretable model and it is able to control overfitting, even on small samples. In the present work we demonstrate through a comprehensive simulation study the superior performance of hiPRS w.r.t. state of the art methods, both in terms of scoring performance and interpretability of the resulting model. We also test hiPRS against small sample size, class imbalance and the presence of noise, showcasing its robustness to extreme experimental settings. Finally, we apply hiPRS to a case study on real data from DACHS cohort, defining an interaction-aware scoring model to predict mortality of stage II-III Colon-Rectal Cancer patients treated with oxaliplatin.

AI in medicine: Where are we now and where are we going?

Advancements in AI enable personalizing healthcare, for example by investigating disease origins at the genetic or molecular level, understanding intraindividual drug effects, and fusing multi-modal personal physiological, behavioral, laboratory, and clinical data to uncover new aspects of pathophysiology. Future efforts should address equity, fairness, explainability, and generalizability of AI models.

Machine learning optimized polygenic scores for blood cell traits identify sex-specific trajectories and genetic correlations with disease

Genetic association studies for blood cell traits, which are key indicators of health and immune function, have identified several hundred associations and defined a complex polygenic architecture. Polygenic scores (PGSs) for blood cell traits have potential clinical utility in disease risk prediction and prevention, but designing PGS remains challenging and the optimal methods are unclear. To address this, we evaluated the relative performance of 6 methods to develop PGS for 26 blood cell traits, including a standard method of pruning and thresholding (P + T) and 5 learning methods: LDpred2, elastic net (EN), Bayesian ridge (BR), multilayer perceptron (MLP) and convolutional neural network (CNN). We evaluated these optimized PGSs on blood cell trait data from UK Biobank and INTERVAL. We find that PGSs designed using common machine learning methods EN and BR show improved prediction of blood cell traits and consistently outperform other methods. Our analyses suggest EN/BR as the top choices for PGS construction, showing improved performance for 25 blood cell traits in the external validation, with correlations with the directly measured traits increasing by 10%-23%. Ten PGSs showed significant statistical interaction with sex, and sex-specific PGS stratification showed that all of them had substantial variation in the trajectories of blood cell traits with age. Genetic correlations between the PGSs for blood cell traits and common human diseases identified well-known as well as new associations. We develop machine learning-optimized PGS for blood cell traits, demonstrate their relationships with sex, age, and disease, and make these publicly available as a resource.

From Genotype to Phenotype: Polygenic Prediction of Complex Human Traits

Decoding the genome confers the capability to predict characteristics of the organism (phenotype) from DNA (genotype). We describe the present status and future prospects of genomic prediction of complex traits in humans. Some highly heritable complex phenotypes such as height and other quantitative traits can already be predicted with reasonable accuracy from DNA alone. For many diseases, including important common conditions such as coronary artery disease, breast cancer, type I and II diabetes, individuals with outlier polygenic scores (e.g., top few percent) have been shown to have 5 or even 10 times higher risk than average. Several psychiatric conditions such as schizophrenia and autism also fall into this category. We discuss related topics such as the genetic architecture of complex traits, sibling validation of polygenic scores, and applications to adult health, in vitro fertilization (embryo selection), and genetic engineering.

Capturing SNP Association across the NK Receptor and HLA Gene Regions in Multiple Sclerosis by Targeted Penalised Regression Models

Conventional genome-wide association studies (GWASs) of complex traits, such as Multiple Sclerosis (MS), are reliant on per-SNP p-values and are therefore heavily burdened by multiple testing correction. Thus, in order to detect more subtle alterations, ever increasing sample sizes are required, while ignoring potentially valuable information that is readily available in existing datasets. To overcome this, we used penalised regression incorporating elastic net with a stability selection method by iterative subsampling to detect the potential interaction of loci with MS risk. Through re-analysis of the ANZgene dataset (1617 cases and 1988 controls) and an IMSGC dataset as a replication cohort (1313 cases and 1458 controls), we identified new association signals for MS predisposition, including SNPs above and below conventional significance thresholds while targeting two natural killer receptor loci and the well-established HLA loci. For example, rs2844482 (98.1% iterations), otherwise ignored by conventional statistics (p = 0.673) in the same dataset, was independently strongly associated with MS in another GWAS that required more than 40 times the number of cases (~45 K). Further comparison of our hits to those present in a large-scale meta-analysis, confirmed that the majority of SNPs identified by the elastic net model reached conventional statistical GWAS thresholds (p < 5 × 10−8) in this much larger dataset. Moreover, we found that gene variants involved in oxidative stress, in addition to innate immunity, were associated with MS. Overall, this study highlights the benefit of using more advanced statistical methods to (re-)analyse subtle genetic variation among loci that have a biological basis for their contribution to disease risk.

Artificial intelligence and diabetes technology: A review

Artificial intelligence (AI) is widely discussed in the popular literature and is portrayed as impacting many aspects of human life, both in and out of the workplace. The potential for revolutionizing healthcare is significant because of the availability of increasingly powerful computational platforms and methods, along with increasingly informative sources of patient data, both in and out of clinical settings. This review aims to provide a realistic assessment of the potential for AI in understanding and managing diabetes, accounting for the state of the art in the methodology and medical devices that collect data, process data, and act accordingly. Acknowledging that many conflicting definitions of AI have been put forth, this article attempts to characterize the main elements of the field as they relate to diabetes, identifying the main perspectives and methods that can (i) affect basic understanding of the disease, (ii) affect understanding of risk factors (genetic, clinical, and behavioral) of diabetes development, (iii) improve diagnosis, (iv) improve understanding of the arc of disease (progression and personal/societal impact), and finally (v) improve treatment.

Machine learning for precision medicine

Precision medicine is an emerging approach to clinical research and patient care that focuses on understanding and treating disease by integrating multi-modal or multi-omics data from an individual to make patient-tailored decisions. With the large and complex datasets generated using precision medicine diagnostic approaches, novel techniques to process and understand these complex data were needed. At the same time, computer science has progressed rapidly to develop techniques that enable the storage, processing, and analysis of these complex datasets, a feat that traditional statistics and early computing technologies could not accomplish. Machine learning, a branch of artificial intelligence, is a computer science methodology that aims to identify complex patterns in data that can be used to make predictions or classifications on new unseen data or for advanced exploratory data analysis. Machine learning analysis of precision medicine's multi-modal data allows for broad analysis of large datasets and ultimately a greater understanding of human health and disease. This review focuses on machine learning utilization for precision medicine's "big data", in the context of genetics, genomics, and beyond.

Deep neural network improves the estimation of polygenic risk scores for breast cancer

Polygenic risk scores (PRS) estimate the genetic risk of an individual for a complex disease based on many genetic variants across the whole genome. In this study, we compared a series of computational models for estimation of breast cancer PRS. A deep neural network (DNN) was found to outperform alternative machine learning techniques and established statistical algorithms, including BLUP, BayesA, and LDpred. In the test cohort with 50% prevalence, the Area Under the receiver operating characteristic Curve (AUC) were 67.4% for DNN, 64.2% for BLUP, 64.5% for BayesA, and 62.4% for LDpred. BLUP, BayesA, and LPpred all generated PRS that followed a normal distribution in the case population. However, the PRS generated by DNN in the case population followed a bimodal distribution composed of two normal distributions with distinctly different means. This suggests that DNN was able to separate the case population into a high-genetic-risk case subpopulation with an average PRS significantly higher than the control population and a normal-genetic-risk case subpopulation with an average PRS similar to the control population. This allowed DNN to achieve 18.8% recall at 90% precision in the test cohort with 50% prevalence, which can be extrapolated to 65.4% recall at 20% precision in a general population with 12% prevalence. Interpretation of the DNN model identified salient variants that were assigned insignificant p values by association studies, but were important for DNN prediction. These variants may be associated with the phenotype through nonlinear relationships.

Penalized regression and model selection methods for polygenic scores on summary statistics

Polygenic scores quantify the genetic risk associated with a given phenotype and are widely used to predict the risk of complex diseases. There has been recent interest in developing methods to construct polygenic risk scores using summary statistic data. We propose a method to construct polygenic risk scores via penalized regression using summary statistic data and publicly available reference data. Our method bears similarity to existing method LassoSum, extending their framework to the Truncated Lasso Penalty (TLP) and the elastic net. We show via simulation and real data application that the TLP improves predictive accuracy as compared to the LASSO while imposing additional sparsity where appropriate. To facilitate model selection in the absence of validation data, we propose methods for estimating model fitting criteria AIC and BIC. These methods approximate the AIC and BIC in the case where we have a polygenic risk score estimated on summary statistic data and no validation data. Additionally, we propose the so-called quasi-correlation metric, which quantifies the predictive accuracy of a polygenic risk score applied to out-of-sample data for which we have only summary statistic information. In total, these methods facilitate estimation and model selection of polygenic risk scores on summary statistic data, and the application of these polygenic risk scores to out-of-sample data for which we have only summary statistic information. We demonstrate the utility of these methods by applying them to GWA studies of lipids, height, and lung cancer.

Mining genetic and transcriptomic data using machine learning approaches in Parkinson's disease

High-throughput techniques have generated abundant genetic and transcriptomic data of Parkinson’s disease (PD) patients but data analysis approaches such as traditional statistical methods have not provided much in the way of insightful integrated analysis or interpretation of the data. As an advanced computational approach, machine learning, which enables people to identify complex patterns and insight from data, has consequently been harnessed to analyze and interpret large, highly complex genetic and transcriptomic data toward a better understanding of PD. In particular, machine learning models have been developed to integrate patient genotype data alone or combined with demographic, clinical, neuroimaging, and other information, for PD outcome study. They have also been used to identify biomarkers of PD based on transcriptomic data, e.g., gene expression profiles from microarrays. This study overviews the relevant literature on using machine learning models for genetic and transcriptomic data analysis in PD, points out remaining challenges, and suggests future directions accordingly. Undoubtedly, the use of machine learning is amplifying PD genetic and transcriptomic achievements for accelerating the study of PD. Existing studies have demonstrated the great potential of machine learning in discovering hidden patterns within genetic or transcriptomic information and thus revealing clues underpinning pathology and pathogenesis. Moving forward, by addressing the remaining challenges, machine learning may advance our ability to precisely diagnose, prognose, and treat PD.

Genome-wide Modeling of Polygenic Risk Score in Colorectal Cancer Risk

Accurate colorectal cancer (CRC) risk prediction models are critical for identifying individuals at low and high risk of developing CRC, as they can then be offered targeted screening and interventions to address their risks of developing disease (if they are in a high-risk group) and avoid unnecessary screening and interventions (if they are in a low-risk group). As it is likely that thousands of genetic variants contribute to CRC risk, it is clinically important to investigate whether these genetic variants can be used jointly for CRC risk prediction. In this paper, we derived and compared different approaches to generating predictive polygenic risk scores (PRS) from genome-wide association studies (GWASs) including 55,105 CRC-affected case subjects and 65,079 control subjects of European ancestry. We built the PRS in three ways, using (1) 140 previously identified and validated CRC loci; (2) SNP selection based on linkage disequilibrium (LD) clumping followed by machine-learning approaches; and (3) LDpred, a Bayesian approach for genome-wide risk prediction. We tested the PRS in an independent cohort of 101,987 individuals with 1,699 CRC-affected case subjects. The discriminatory accuracy, calculated by the age- and sex-adjusted area under the receiver operating characteristics curve (AUC), was highest for the LDpred-derived PRS (AUC = 0.654) including nearly 1.2 M genetic variants (the proportion of causal genetic variants for CRC assumed to be 0.003), whereas the PRS of the 140 known variants identified from GWASs had the lowest AUC (AUC = 0.629). Based on the LDpred-derived PRS, we are able to identify 30% of individuals without a family history as having risk for CRC similar to those with a family history of CRC, whereas the PRS based on known GWAS variants identified only top 10% as having a similar relative risk. About 90% of these individuals have no family history and would have been considered average risk under current screening guidelines, but might benefit from earlier screening. The developed PRS offers a way for risk-stratified CRC screening and other targeted interventions.

SummaryAUC: a tool for evaluating the performance of polygenic risk prediction models in validation datasets with only summary level statistics

MOTIVATION

Polygenic risk score (PRS) methods based on genome-wide association studies (GWAS) have a potential for predicting the risk of developing complex diseases and are expected to become more accurate with larger training datasets and innovative statistical methods. The area under the ROC curve (AUC) is often used to evaluate the performance of PRSs, which requires individual genotypic and phenotypic data in an independent GWAS validation dataset. We are motivated to develop methods for approximating AUC of PRSs based on the summary level data of the validation dataset, which will greatly facilitate the development of PRS models for complex diseases.

RESULTS

We develop statistical methods and an R package SummaryAUC for approximating the AUC and its variance of a PRS when only the summary level data of the validation dataset are available. SummaryAUC can be applied to PRSs with SNPs either genotyped or imputed in the validation dataset. We examined the performance of SummaryAUC using a large-scale GWAS of schizophrenia. SummaryAUC provides accurate approximations to AUCs and their variances. The bias of AUC is typically <0.5% in most analyses. SummaryAUC cannot be applied to PRSs that use all SNPs in the genome because it is computationally prohibitive.

AVAILABILITY AND IMPLEMENTATION

https://github.com/lsncibb/SummaryAUC.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Predicting Phenotypic Polymyxin Resistance in Klebsiella pneumoniae through Machine Learning Analysis of Genomic Data

Polymyxins are used as treatments of last resort for Gram-negative bacterial infections. Their increased use has led to concerns about emerging polymyxin resistance (PR). Phenotypic polymyxin susceptibility testing is resource intensive and difficult to perform accurately. The complex polygenic nature of PR and our incomplete understanding of its genetic basis make it difficult to predict PR using detection of resistance determinants. We therefore applied machine learning (ML) to whole-genome sequencing data from >600 Klebsiella pneumoniae clonal group 258 (CG258) genomes to predict phenotypic PR. Using a reference-based representation of genomic data with ML outperformed a rule-based approach that detected variants in known PR genes (area under receiver-operator curve [AUROC], 0.894 versus 0.791, P = 0.006). We noted modest increases in performance by using a bacterial genome-wide association study to filter relevant genomic features and by integrating clinical data in the form of prior polymyxin exposure. Conversely, reference-free representation of genomic data as k-mers was associated with decreased performance (AUROC, 0.692 versus 0.894, P = 0.015). When ML models were interpreted to extract genomic features, six of seven known PR genes were correctly identified by models without prior programming and several genes involved in stress responses and maintenance of the cell membrane were identified as potential novel determinants of PR. These findings are a proof of concept that whole-genome sequencing data can accurately predict PR in K. pneumoniae CG258 and may be applicable to other forms of complex antimicrobial resistance.IMPORTANCE Polymyxins are last-resort antibiotics used to treat highly resistant Gram-negative bacteria. There are increasing reports of polymyxin resistance emerging, raising concerns of a postantibiotic era. Polymyxin resistance is therefore a significant public health threat, but current phenotypic methods for detection are difficult and time-consuming to perform. There have been increasing efforts to use whole-genome sequencing for detection of antibiotic resistance, but this has been difficult to apply to polymyxin resistance because of its complex polygenic nature. The significance of our research is that we successfully applied machine learning methods to predict polymyxin resistance in Klebsiella pneumoniae clonal group 258, a common health care-associated and multidrug-resistant pathogen. Our findings highlight that machine learning can be successfully applied even in complex forms of antibiotic resistance and represent a significant contribution to the literature that could be used to predict resistance in other bacteria and to other antibiotics.

Improving predictive models for Alzheimer's disease using GWAS data by incorporating misclassified samples modeling

Late-onset Alzheimer’s Disease (LOAD) is the most common form of dementia in the elderly. Genome-wide association studies (GWAS) for LOAD have open new avenues to identify genetic causes and to provide diagnostic tools for early detection. Although several predictive models have been proposed using the few detected GWAS markers, there is still a need for improvement and identification of potential markers. Commonly, polygenic risk scores are being used for prediction. Nevertheless, other methods to generate predictive models have been suggested. In this research, we compared three machine learning methods that have been proved to construct powerful predictive models (genetic algorithms, LASSO, and step-wise) and propose the inclusion of markers from misclassified samples to improve overall prediction accuracy. Our results show that the addition of markers from an initial model plus the markers of the model fitted to misclassified samples improves the area under the receiving operative curve by around 5%, reaching ~0.84, which is highly competitive using only genetic information. The computational strategy used here can help to devise better methods to improve classification models for AD. Our results could have a positive impact on the early diagnosis of Alzheimer’s disease.

A systematic review of the applications of artificial intelligence and machine learning in autoimmune diseases

Autoimmune diseases are chronic, multifactorial conditions. Through machine learning (ML), a branch of the wider field of artificial intelligence, it is possible to extract patterns within patient data, and exploit these patterns to predict patient outcomes for improved clinical management. Here, we surveyed the use of ML methods to address clinical problems in autoimmune disease. A systematic review was conducted using MEDLINE, embase and computers and applied sciences complete databases. Relevant papers included "machine learning" or "artificial intelligence" and the autoimmune diseases search term(s) in their title, abstract or key words. Exclusion criteria: studies not written in English, no real human patient data included, publication prior to 2001, studies that were not peer reviewed, non-autoimmune disease comorbidity research and review papers. 169 (of 702) studies met the criteria for inclusion. Support vector machines and random forests were the most popular ML methods used. ML models using data on multiple sclerosis, rheumatoid arthritis and inflammatory bowel disease were most common. A small proportion of studies (7.7% or 13/169) combined different data types in the modelling process. Cross-validation, combined with a separate testing set for more robust model evaluation occurred in 8.3% of papers (14/169). The field may benefit from adopting a best practice of validation, cross-validation and independent testing of ML models. Many models achieved good predictive results in simple scenarios (e.g. classification of cases and controls). Progression to more complex predictive models may be achievable in future through integration of multiple data types.

Systematic Review of Polygenic Risk Scores for Type 1 and Type 2 Diabetes

Recent studies have led to considerable advances in the identification of genetic variants associated with type 1 and type 2 diabetes. An approach for converting genetic data into a predictive measure of disease susceptibility is to add the risk effects of loci into a polygenic risk score. In order to summarize the recent findings, we conducted a systematic review of studies comparing the accuracy of polygenic risk scores developed during the last two decades. We selected 15 risk scores from three databases (Scopus, Web of Science and PubMed) enrolled in this systematic review. We identified three polygenic risk scores that discriminate between type 1 diabetes patients and healthy people, one that discriminate between type 1 and type 2 diabetes, two that discriminate between type 1 and monogenic diabetes and nine polygenic risk scores that discriminate between type 2 diabetes patients and healthy people. Prediction accuracy of polygenic risk scores was assessed by comparing the area under the curve. The actual benefits, potential obstacles and possible solutions for the implementation of polygenic risk scores in clinical practice were also discussed. Develop strategies to establish the clinical validity of polygenic risk scores by creating a framework for the interpretation of findings and their translation into actual evidence, are the way to demonstrate their utility in medical practice.

Machine Learning SNP Based Prediction for Precision Medicine

In the past decade, precision genomics based medicine has emerged to provide tailored and effective healthcare for patients depending upon their genetic features. Genome Wide Association Studies have also identified population based risk genetic variants for common and complex diseases. In order to meet the full promise of precision medicine, research is attempting to leverage our increasing genomic understanding and further develop personalized medical healthcare through ever more accurate disease risk prediction models. Polygenic risk scoring and machine learning are two primary approaches for disease risk prediction. Despite recent improvements, the results of polygenic risk scoring remain limited due to the approaches that are currently used. By contrast, machine learning algorithms have increased predictive abilities for complex disease risk. This increase in predictive abilities results from the ability of machine learning algorithms to handle multi-dimensional data. Here, we provide an overview of polygenic risk scoring and machine learning in complex disease risk prediction. We highlight recent machine learning application developments and describe how machine learning approaches can lead to improved complex disease prediction, which will help to incorporate genetic features into future personalized healthcare. Finally, we discuss how the future application of machine learning prediction models might help manage complex disease by providing tissue-specific targets for customized, preventive interventions.

Efficient Implementation of Penalized Regression for Genetic Risk Prediction

Polygenic risk scores (PRS) combine many single-nucleotide polymorphisms into a score reflecting the genetic risk of developing a disease. Privé, Aschard, and Blum present an efficient implementation of penalized logistic regression...

Polygenic Risk Scores (PRS) combine genotype information across many single-nucleotide polymorphisms (SNPs) to give a score reflecting the genetic risk of developing a disease. PRS might have a major impact on public health, possibly allowing for screening campaigns to identify high-genetic risk individuals for a given disease. The “Clumping+Thresholding” (C+T) approach is the most common method to derive PRS. C+T uses only univariate genome-wide association studies (GWAS) summary statistics, which makes it fast and easy to use. However, previous work showed that jointly estimating SNP effects for computing PRS has the potential to significantly improve the predictive performance of PRS as compared to C+T. In this paper, we present an efficient method for the joint estimation of SNP effects using individual-level data, allowing for practical application of penalized logistic regression (PLR) on modern datasets including hundreds of thousands of individuals. Moreover, our implementation of PLR directly includes automatic choices for hyper-parameters. We also provide an implementation of penalized linear regression for quantitative traits. We compare the performance of PLR, C+T and a derivation of random forests using both real and simulated data. Overall, we find that PLR achieves equal or higher predictive performance than C+T in most scenarios considered, while being scalable to biobank data. In particular, we find that improvement in predictive performance is more pronounced when there are few effects located in nearby genomic regions with correlated SNPs; for instance, in simulations, AUC values increase from 83% with the best prediction of C+T to 92.5% with PLR. We confirm these results in a data analysis of a case-control study for celiac disease where PLR and the standard C+T method achieve AUC values of 89% and of 82.5%. Applying penalized linear regression to 350,000 individuals of the UK Biobank, we predict height with a larger correlation than with the best prediction of C+T (∼65% instead of ∼55%), further demonstrating its scalability and strong predictive power, even for highly polygenic traits. Moreover, using 150,000 individuals of the UK Biobank, we are able to predict breast cancer better than C+T, fitting PLR in a few minutes only. In conclusion, this paper demonstrates the feasibility and relevance of using penalized regression for PRS computation when large individual-level datasets are available, thanks to the efficient implementation available in our R package bigstatsr.

A Pharmacogenetic Prediction Model of Progression-Free Survival in Breast Cancer using Genome-Wide Genotyping Data from CALGB 40502 (Alliance)

Genome-wide genotyping data are increasingly available for pharmacogenetic association studies, but application of these data for development of prediction models is limited. Prediction methods, such as elastic net regularization, have recently been applied to genetic studies but only limitedly to pharmacogenetic outcomes. An elastic net was applied to a pharmacogenetic study of progression-free survival (PFS) of 468 patients with advanced breast cancer in a clinical trial of paclitaxel, nab-paclitaxel, and ixabepilone. A final model included 13 single nucleotide polymorphisms (SNPs) in addition to clinical covariates (prior taxane status, hormone receptor status, disease-free interval, and presence of visceral metastases) with an area under the curve (AUC) integrated over time of 0.81, an increase compared to an AUC of 0.64 for a model with clinical covariates alone. This model may be of value in predicting PFS with microtubule targeting agents and may inform reverse translational studies to understand differential response to these drugs.

Using Full Genomic Information to Predict Disease: Breaking Down the Barriers Between Complex and Mendelian Diseases

While sequence-based genetic tests have long been available for specific loci, especially for Mendelian disease, the rapidly falling costs of genome-wide genotyping arrays, whole-exome sequencing, and whole-genome sequencing are moving us toward a future where full genomic information might inform the prognosis and treatment of a variety of diseases, including complex disease. Similarly, the availability of large populations with full genomic information has enabled new insights about the etiology and genetic architecture of complex disease. Insights from the latest generation of genomic studies suggest that our categorization of diseases as complex may conceal a wide spectrum of genetic architectures and causal mechanisms that ranges from Mendelian forms of complex disease to complex regulatory structures underlying Mendelian disease. Here, we review these insights, along with advances in the prediction of disease risk and outcomes from full genomic information.

GWAS-based machine learning approach to predict duloxetine response in major depressive disorder

Major depressive disorder (MDD) is one of the most prevalent psychiatric disorders and is commonly treated with antidepressant drugs. However, large variability is observed in terms of response to antidepressants. Machine learning (ML) models may be useful to predict treatment outcomes. A sample of 186 MDD patients received treatment with duloxetine for up to 8 weeks were categorized as "responders" based on a MADRS change >50% from baseline; or "remitters" based on a MADRS score ≤10 at end point. The initial dataset (N = 186) was randomly divided into training and test sets in a nested 5-fold cross-validation, where 80% was used as a training set and 20% made up five independent test sets. We performed genome-wide logistic regression to identify potentially significant variants related to duloxetine response/remission and extracted the most promising predictors using LASSO regression. Subsequently, classification-regression trees (CRT) and support vector machines (SVM) were applied to construct models, using ten-fold cross-validation. With regards to response, none of the pairs performed significantly better than chance (accuracy p > .1). For remission, SVM achieved moderate performance with an accuracy = 0.52, a sensitivity = 0.58, and a specificity = 0.46, and 0.51 for all coefficients for CRT. The best performing SVM fold was characterized by an accuracy = 0.66 (p = .071), sensitivity = 0.70 and a sensitivity = 0.61. In this study, the potential of using GWAS data to predict duloxetine outcomes was examined using ML models. The models were characterized by a promising sensitivity, but specificity remained moderate at best. The inclusion of additional non-genetic variables to create integrated models may improve prediction.

Rationale and design of GENEiUS: a prospective observational study on the genetic and environmental determinants of body mass index evolution in Canadian undergraduate students

INTRODUCTION

Obesity is a global epidemic and is a risk factor for developing other comorbidities. Young adulthood is a critical period for body weight change and establishing healthy lifestyle behaviours. The 'Freshman 15' suggests that undergraduate students gain 15 lbs (6.8 kg) during their first year of university, although evidence estimates a more modest weight gain of approximately 3-5 lbs (1.4-2.3 kg). Previous studies have only investigated weight change in the first year and do not study potential risk factors. Genetic and EnviroNmental Effects on weight in University Students (GENEiUS) is a prospective observational study which will investigate the environmental and biological determinants of weight change in undergraduate students over 4 years.

METHODS AND ANALYSIS

The GENEiUS study will recruit 2500 multiethnic undergraduates aged 17-25 years at McMaster University at the start of their first year and will follow them every 6 months for 4 years. Primary outcomes are obesity traits: body mass index, waist circumference, waist-to-hip ratio, body fat mass and body fat percentage. The contribution of well-established and novel genetic variants for obesity traits and heritability values will be derived from whole-genome single-nucleotide polymorphism genotyping arrays. Civil status, age, sex, ethnicity, length of residence in Canada, religiosity, energy intake, physical activity, exercise motivation, electronic screen time, sleep patterns, history of assault, smoking status, alcohol consumption, medication and drug use, stress, impulsivity, body image perception, self-esteem, anxiety, eating disorders and depression will be investigated for their effect on obesity traits. The findings of the GENEiUS study will be used to help design obesity prevention programme in North American universities with multiethnic populations.

ETHICS AND DISSEMINATION

Ethical approval of the study protocol has been obtained from the Hamilton Integrated Research Ethics Board. Study results will be disseminated through scientific publications, scholarly meetings, and collaborative meetings with university administration and student groups.

Improving Disease Prediction by Incorporating Family Disease History in Risk Prediction Models with Large-Scale Genetic Data

Despite the many successes of genome-wide association studies (GWAS), the known susceptibility variants identified by GWAS have modest effect sizes, leading to notable skepticism about the effectiveness of building a risk prediction model from large-scale genetic data. However, in contrast to genetic variants, the family history of diseases has been largely accepted as an important risk factor in clinical diagnosis and risk prediction. Nevertheless, the complicated structures of the family history of diseases have limited their application in clinical practice. Here, we developed a new method that enables incorporation of the general family history of diseases with a liability threshold model, and propose a new analysis strategy for risk prediction with penalized regression analysis that incorporates both large numbers of genetic variants and clinical risk factors. Application of our model to type 2 diabetes in the Korean population (1846 cases and 1846 controls) demonstrated that single-nucleotide polymorphisms accounted for 32.5% of the variation explained by the predicted risk scores in the test data set, and incorporation of family history led to an additional 6.3% improvement in prediction. Our results illustrate that family medical history provides valuable information on the variation of complex diseases and improves prediction performance.

Assessing the Heritability of Complex Traits in Humans: Methodological Challenges and Opportunities

The goal of this review article is to provide a conceptual based summary of how heritability estimates for complex traits such as obesity are determined and to explore the future directions of research in the heritability field. The target audience are researchers who use heritability data rather than those conducting heritability studies. The article provides an introduction to key concepts critical to understanding heritability studies including: i) definitions of heritability: broad sense versus narrow sense heritability; ii) how data for heritability studies are collected: twin, adoption, family and population-based studies; and iii) analytical techniques: path analysis, structural equations and mixed or regressive models of complex segregation analysis. For each section, a discussion of how the different definitions and methodologies influence heritability estimates is provided. The general limitations of heritability studies are discussed including the issue of “missing heritability” in which heritability estimates are significantly higher than the variance explained by known genetic variants. Potential causes of missing heritability include restriction of many genetic association studies to single nucleotide polymorphisms, gene by gene interactions, epigenetics, and gene by environment interactions. Innovative strategies of accounting for missing heritability including modeling techniques and improved software are discussed.

Association of HLA Genetic Risk Burden With Disease Phenotypes in Multiple Sclerosis

IMPORTANCE

Although multiple HLA alleles associated with multiple sclerosis (MS) risk have been identified, genotype-phenotype studies in the HLA region remain scarce and inconclusive.

OBJECTIVES

To investigate whether MS risk-associated HLA alleles also affect disease phenotypes.

DESIGN, SETTING, AND PARTICIPANTS

A cross-sectional, case-control study comprising 652 patients with MS who had comprehensive phenotypic information and 455 individuals of European origin serving as controls was conducted at a single academic research site. Patients evaluated at the Multiple Sclerosis Center at University of California, San Francisco between July 2004 and September 2005 were invited to participate. Spinal cord imaging in the data set was acquired between July 2013 and March 2014; analysis was performed between December 2014 and December 2015.

MAIN OUTCOMES AND MEASURES

Cumulative HLA genetic burden (HLAGB) calculated using the most updated MS-associated HLA alleles vs clinical and magnetic resonance imaging outcomes, including age at onset, disease severity, conversion time from clinically isolated syndrome to clinically definite MS, fractions of cortical and subcortical gray matter and cerebral white matter, brain lesion volume, spinal cord gray and white matter areas, upper cervical cord area, and the ratio of gray matter to the upper cervical cord area. Multivariate modeling was applied separately for each sex data set.

RESULTS

Of the 652 patients with MS, 586 had no missing genetic data and were included in the HLAGB analysis. In these 586 patients (404 women [68.9%]; mean [SD] age at disease onset, 33.6 [9.4] years), HLAGB was higher than in controls (median [IQR], 0.7 [0-1.4] and 0 [-0.3 to 0.5], respectively; P = 1.8 × 10-27). A total of 619 (95.8%) had relapsing-onset MS and 27 (4.2%) had progressive-onset MS. No significant difference was observed between relapsing-onset MS and primary progressive MS. A higher HLAGB was associated with younger age at onset and the atrophy of subcortical gray matter fraction in women with relapsing-onset MS (standard β = -1.20 × 10-1; P = 1.7 × 10-2 and standard β = -1.67 × 10-1; P = 2.3 × 10-4, respectively), which were driven mainly by the HLA-DRB1*15:01 haplotype. In addition, we observed the distinct role of the HLA-A*24:02-B*07:02-DRB1*15:01 haplotype among the other common DRB1*15:01 haplotypes and a nominally protective effect of HLA-B*44:02 to the subcortical gray atrophy (standard β = -1.28 × 10-1; P = 5.1 × 10-3 and standard β = 9.52 × 10-2; P = 3.6 × 10-2, respectively).

CONCLUSIONS AND RELEVANCE

We confirm and extend previous observations linking HLA MS susceptibility alleles with disease progression and specific clinical and magnetic resonance imaging phenotypic traits.

Sibling method increases risk assessment estimates for type 1 diabetes

We presented a risk assessment model to distinguish between type 1 diabetes (T1D) affected and unaffected siblings using only three single nucleotide polymorphism (SNP) genotypes. In addition we calculated the heritability from genome-wide identity-by-descent (IBD) sharing between full siblings. We analyzed 1,253 pairs of affected individuals and their unaffected siblings (750 pairs from a discovery set and 503 pairs from a validation set) from the T1D Genetics Consortium (T1DGC), applying a logistic regression to analyze the area under the receiver operator characteristic (ROC) curve (AUC). To calculate the heritability of T1D we used the Haseman-Elston regression analysis of the squared difference between the phenotypes of the pairs of siblings on the estimate of their genome-wide IBD proportion. The model with only 3 SNPs achieving an AUC of 0.75 in both datasets outperformed the model using the presence of the high-risk DR3/4 HLA genotype, namely AUC of 0.60. The heritability on the liability scale of T1D was approximately from 0.53 to 0.92, close to the results obtained from twin studies, ranging from 0.4 to 0.88.

Recent progress in genetics, epigenetics and metagenomics unveils the pathophysiology of human obesity

In high-, middle- and low-income countries, the rising prevalence of obesity is the underlying cause of numerous health complications and increased mortality. Being a complex and heritable disorder, obesity results from the interplay between genetic susceptibility, epigenetics, metagenomics and the environment. Attempts at understanding the genetic basis of obesity have identified numerous genes associated with syndromic monogenic, non-syndromic monogenic, oligogenic and polygenic obesity. The genetics of leanness are also considered relevant as it mirrors some of obesity's aetiologies. In this report, we summarize ten genetically elucidated obesity syndromes, some of which are involved in ciliary functioning. We comprehensively review 11 monogenic obesity genes identified to date and their role in energy maintenance as part of the leptin-melanocortin pathway. With the emergence of genome-wide association studies over the last decade, 227 genetic variants involved in different biological pathways (central nervous system, food sensing and digestion, adipocyte differentiation, insulin signalling, lipid metabolism, muscle and liver biology, gut microbiota) have been associated with polygenic obesity. Advances in obligatory and facilitated epigenetic variation, and gene-environment interaction studies have partly accounted for the missing heritability of obesity and provided additional insight into its aetiology. The role of gut microbiota in obesity pathophysiology, as well as the 12 genes associated with lipodystrophies is discussed. Furthermore, in an attempt to improve future studies and merge the gap between research and clinical practice, we provide suggestions on how high-throughput '-omic' data can be integrated in order to get closer to the new age of personalized medicine.

Using information of relatives in genomic prediction to apply effective stratified medicine

Genomic prediction shows promise for personalised medicine in which diagnosis and treatment are tailored to individuals based on their genetic profiles for complex diseases. We present a theoretical framework to demonstrate that prediction accuracy can be improved by targeting more informative individuals in the data set used to generate the predictors ("discovery sample") to include those with genetically close relationships with the subjects put forward for risk prediction. Increase of prediction accuracy from closer relationships is achieved under an additive model and does not rely on any family or interaction effects. Using theory, simulations and real data analyses, we show that the predictive accuracy or the area under the receiver operating characteristic curve (AUC) increased exponentially with decreasing effective size (Ne), i.e. when individuals are closely related. For example, with the sample size of discovery set N = 3000, heritability h2 = 0.5 and population prevalence K = 0.1, AUC value approached to 0.9 and the top percentile of the estimated genetic profile scores had 23 times higher proportion of cases than the general population. This suggests that there is considerable room to increase prediction accuracy by using a design that does not exclude closer relationships.

Collective Genetic Interaction Effects and the Role of Antigen-Presenting Cells in Autoimmune Diseases

Autoimmune diseases occur when immune cells fail to develop or lose their tolerance toward self and destroy body's own tissues. Both insufficient negative selection of self-reactive T cells and impaired development of regulatory T cells preventing effector cell activation are believed to contribute to autoimmunity. Genetic predispositions center around the major histocompatibility complex (MHC) class II loci involved in antigen presentation, the key determinant of CD4+ T cell activation. Recent studies suggested that variants in the MHC region also exhibit significant non-additive interaction effects. However, collective interactions involving large numbers of single nucleotide polymorphisms (SNPs) contributing to such effects are yet to be characterized. In addition, relatively little is known about the cell-type-specificity of such interactions in the context of cellular pathways. Here, we analyzed type 1 diabetes (T1D) and rheumatoid arthritis (RA) genome-wide association data sets via large-scale, high-performance computations and inferred collective interaction effects involving MHC SNPs using the discrete discriminant analysis. Despite considerable differences in the details of SNP interactions in T1D and RA data, the enrichment pattern of interacting pairs in reference epigenomes was remarkably similar: statistically significant interactions were epigenetically active in cell-type combinations connecting B cells to T cells and intestinal epithelial cells, with both helper and regulatory T cells showing strong disease-associated interactions with B cells. Our results provide direct genetic evidence pointing to the important roles B cells play as antigen-presenting cells toward CD4+ T cells in the context of central and peripheral tolerance. In addition, they are consistent with recent experimental studies suggesting that the repertoire of B cell-specific self-antigens in the thymus are critical to the effective control of corresponding autoimmune activation in peripheral tissues.

Combining Multiple Hypothesis Testing with Machine Learning Increases the Statistical Power of Genome-wide Association Studies

The standard approach to the analysis of genome-wide association studies (GWAS) is based on testing each position in the genome individually for statistical significance of its association with the phenotype under investigation. To improve the analysis of GWAS, we propose a combination of machine learning and statistical testing that takes correlation structures within the set of SNPs under investigation in a mathematically well-controlled manner into account. The novel two-step algorithm, COMBI, first trains a support vector machine to determine a subset of candidate SNPs and then performs hypothesis tests for these SNPs together with an adequate threshold correction. Applying COMBI to data from a WTCCC study (2007) and measuring performance as replication by independent GWAS published within the 2008-2015 period, we show that our method outperforms ordinary raw p-value thresholding as well as other state-of-the-art methods. COMBI presents higher power and precision than the examined alternatives while yielding fewer false (i.e. non-replicated) and more true (i.e. replicated) discoveries when its results are validated on later GWAS studies. More than 80% of the discoveries made by COMBI upon WTCCC data have been validated by independent studies. Implementations of the COMBI method are available as a part of the GWASpi toolbox 2.0.

Dissection of the genetic architecture of rice resistance to the blast fungus Magnaporthe oryzae

Resistance in rice cultivars to the rice blast fungus Magnaporthe oryzae is complex and is controlled by both major genes and quantitative trait loci (QTLs). We undertook a genome-wide association study (GWAS) using the rice diversity panel 1 (RDP1) that was genotyped using a high-density (700 000 single nucleotide polymorphisms) array and inoculated with five diverse M. oryzae isolates. We identified 97 loci associated with blast resistance (LABRs). Among them, 82 were new regions and 15 co-localized with known blast resistance loci. The top 72 LABRs explained up to 98% of the phenotypic variation. The candidate genes in the LABRs encode nucleotide-binding site leucine-rich repeat (NBS-LRR) resistance proteins, receptor-like protein kinases, transcription factors and defence-related proteins. Among them, LABR_64 was strongly associated with resistance to all five isolates. We analysed the function of candidate genes underlying LABR_64 using RNA interference (RNAi) technology and identified two new resistance alleles at the Pi5 locus. We demonstrate an efficient strategy for rapid allele discovery using the power of GWAS, coupled with RNAi technology, for the dissection of complex blast resistance in rice.

Development of risk prediction models for glioma based on genome-wide association study findings and comprehensive evaluation of predictive performances

Over 14 common single nucleotide polymorphisms (SNP) have been consistently identified from genome-wide association studies (GWAS) as associated with glioma risk in European background. The extent to which and how these genetic variants can improve the prediction of glioma risk has was not been investigated. In this study, we employed three independent case-control datasets in Chinese populations, tested GWAS signals in dataset1, validated association results in dataset2, developed prediction models in dataset2 for the consistently replicated SNPs, refined the consistently replicated SNPs in dataset3 and developed tailored models for Chinese populations. For model construction, we aggregated the contribution of multiple SNPs into genetic risk scores (count GRS and weighed GRS) or predicted risks from logistic regression analyses (PRFLR).

In dataset2, the area under receiver operating characteristic curves (AUC) of the 5 consistently replicated SNPs by PRFLR(SNPs) was 0.615, higher than those of all GRSs(ranging from 0.607 to 0.611, all P>0.05). The AUC of genetic profile significantly exceeded that of family history (fmc) alone (AUC=0.535, all P<0.001). The best model in our study comprised “PRURA +fmc” (AUC=0.646) in dataset3. Further model assessment analyses provided additional evidence.

This study indicates that genetic markers have potential value for risk prediction of glioma.

Machine learning derived risk prediction of anorexia nervosa

Background

Anorexia nervosa (AN) is a complex psychiatric disease with a moderate to strong genetic contribution. In addition to conventional genome wide association (GWA) studies, researchers have been using machine learning methods in conjunction with genomic data to predict risk of diseases in which genetics play an important role.

Methods

In this study, we collected whole genome genotyping data on 3940 AN cases and 9266 controls from the Genetic Consortium for Anorexia Nervosa (GCAN), the Wellcome Trust Case Control Consortium 3 (WTCCC3), Price Foundation Collaborative Group and the Children’s Hospital of Philadelphia (CHOP), and applied machine learning methods for predicting AN disease risk. The prediction performance is measured by area under the receiver operating characteristic curve (AUC), indicating how well the model distinguishes cases from unaffected control subjects.

Results

Logistic regression model with the lasso penalty technique generated an AUC of 0.693, while Support Vector Machines and Gradient Boosted Trees reached AUC’s of 0.691 and 0.623, respectively. Using different sample sizes, our results suggest that larger datasets are required to optimize the machine learning models and achieve higher AUC values.

Conclusions

To our knowledge, this is the first attempt to assess AN risk based on genome wide genotype level data. Future integration of genomic, environmental and family-based information is likely to improve the AN risk evaluation process, eventually benefitting AN patients and families in the clinical setting.

Electronic supplementary material

The online version of this article (doi:10.1186/s12920-016-0165-x) contains supplementary material, which is available to authorized users.

A comparison of genomic profiles of complex diseases under different models

Background

Various approaches are being used to predict individual risk to polygenic diseases from data provided by genome-wide association studies. As there are substantial differences between the diseases investigated, the data sets used and the way they are tested, it is difficult to assess which models are more suitable for this task.

Results

We compared different approaches for seven complex diseases provided by the Wellcome Trust Case Control Consortium (WTCCC) under a within-study validation approach. Risk models were inferred using a variety of learning machines and assumptions about the underlying genetic model, including a haplotype-based approach with different haplotype lengths and different thresholds in association levels to choose loci as part of the predictive model. In accordance with previous work, our results generally showed low accuracy considering disease heritability and population prevalence. However, the boosting algorithm returned a predictive area under the ROC curve (AUC) of 0.8805 for Type 1 diabetes (T1D) and 0.8087 for rheumatoid arthritis, both clearly over the AUC obtained by other approaches and over 0.75, which is the minimum required for a disease to be successfully tested on a sample at risk, which means that boosting is a promising approach. Its good performance seems to be related to its robustness to redundant data, as in the case of genome-wide data sets due to linkage disequilibrium.

Conclusions

In view of our results, the boosting approach may be suitable for modeling individual predisposition to Type 1 diabetes and rheumatoid arthritis based on genome-wide data and should be considered for more in-depth research.

Electronic supplementary material

The online version of this article (doi:10.1186/s12920-015-0157-2) contains supplementary material, which is available to authorized users.

Progress in understanding type 1 diabetes through its genetic overlap with other autoimmune diseases

Type 1 diabetes mellitus (T1DM) is the most common autoimmune disease in pediatrics with a prevalence of roughly 1 in 500 children in the USA. Genome-wide association studies have identified more than 50 variants associated with increased risk for type 1 diabetes. Comparison of these variants with those identified in other autoimmune diseases reveals three important findings: (1) there is a high degree of overlap in implicated variants in diseases with similar pathophysiology, (2) in diseases with differing pathophysiology the same variants are often implicated in opposite roles, (3) in diseases with differing pathophysiology that have many non-overlapping or oppositely implicated variants there are still several variants which are overlapping or shared. Thus, the genetic overlap between T1DM and other autoimmune diseases forms the basis for our understanding of druggable targets in type 1 diabetes.

Modeling Linkage Disequilibrium Increases Accuracy of Polygenic Risk Scores

Polygenic risk scores have shown great promise in predicting complex disease risk and will become more accurate as training sample sizes increase. The standard approach for calculating risk scores involves linkage disequilibrium (LD)-based marker pruning and applying a p value threshold to association statistics, but this discards information and can reduce predictive accuracy. We introduce LDpred, a method that infers the posterior mean effect size of each marker by using a prior on effect sizes and LD information from an external reference panel. Theory and simulations show that LDpred outperforms the approach of pruning followed by thresholding, particularly at large sample sizes. Accordingly, predicted R(2) increased from 20.1% to 25.3% in a large schizophrenia dataset and from 9.8% to 12.0% in a large multiple sclerosis dataset. A similar relative improvement in accuracy was observed for three additional large disease datasets and for non-European schizophrenia samples. The advantage of LDpred over existing methods will grow as sample sizes increase.

Genetic architecture of colorectal cancer

Colorectal cancer (CRC) is a complex disease that develops as a consequence of both genetic and environmental risk factors. A small proportion (3-5%) of cases arise from hereditary syndromes predisposing to early onset CRC as a result of mutations in over a dozen well defined genes. In contrast, CRC is predominantly a late onset 'sporadic' disease, developing in individuals with no obvious hereditary syndrome. In recent years, genome wide association studies have discovered that over 40 genetic regions are associated with weak effects on sporadic CRC, and it has been estimated that increasingly large genome wide scans will identify many additional novel genetic regions. Subsequent experimental validations have identified the causally related variant(s) in a limited number of these genetic regions. Further biological insight could be obtained through ethnically diverse study populations, larger genetic sequencing studies and development of higher throughput functional experiments. Along with inherited variation, integration of the tumour genome may shed light on the carcinogenic processes in CRC. In addition to summarising the genetic architecture of CRC, this review discusses genetic factors that modify environmental predictors of CRC, as well as examples of how genetic insight has improved clinical surveillance, prevention and treatment strategies. In summary, substantial progress has been made in uncovering the genetic architecture of CRC, and continued research efforts are expected to identify additional genetic risk factors that further our biological understanding of this disease. Subsequently these new insights will lead to improved treatment and prevention of colorectal cancer.

Explicit Modeling of Ancestry Improves Polygenic Risk Scores and BLUP Prediction

Polygenic prediction using genome-wide SNPs can provide high prediction accuracy for complex traits. Here, we investigate the question of how to account for genetic ancestry when conducting polygenic prediction. We show that the accuracy of polygenic prediction in structured populations may be partly due to genetic ancestry. However, we hypothesized that explicitly modeling ancestry could improve polygenic prediction accuracy. We analyzed three GWAS of hair color (HC), tanning ability (TA), and basal cell carcinoma (BCC) in European Americans (sample size from 7,440 to 9,822) and considered two widely used polygenic prediction approaches: polygenic risk scores (PRSs) and best linear unbiased prediction (BLUP). We compared polygenic prediction without correction for ancestry to polygenic prediction with ancestry as a separate component in the model. In 10-fold cross-validation using the PRS approach, the R(2) for HC increased by 66% (0.0456-0.0755; P < 10(-16)), the R(2) for TA increased by 123% (0.0154 to 0.0344; P < 10(-16)), and the liability-scale R(2) for BCC increased by 68% (0.0138-0.0232; P < 10(-16)) when explicitly modeling ancestry, which prevents ancestry effects from entering into each SNP effect and being overweighted. Surprisingly, explicitly modeling ancestry produces a similar improvement when using the BLUP approach, which fits all SNPs simultaneously in a single variance component and causes ancestry to be underweighted. We validate our findings via simulations, which show that the differences in prediction accuracy will increase in magnitude as sample sizes increase. In summary, our results show that explicitly modeling ancestry can be important in both PRS and BLUP prediction.

Influence of Feature Encoding and Choice of Classifier on Disease Risk Prediction in Genome-Wide Association Studies

Various attempts have been made to predict the individual disease risk based on genotype data from genome-wide association studies (GWAS). However, most studies only investigated one or two classification algorithms and feature encoding schemes. In this study, we applied seven different classification algorithms on GWAS case-control data sets for seven different diseases to create models for disease risk prediction. Further, we used three different encoding schemes for the genotypes of single nucleotide polymorphisms (SNPs) and investigated their influence on the predictive performance of these models. Our study suggests that an additive encoding of the SNP data should be the preferred encoding scheme, as it proved to yield the best predictive performances for all algorithms and data sets. Furthermore, our results showed that the differences between most state-of-the-art classification algorithms are not statistically significant. Consequently, we recommend to prefer algorithms with simple models like the linear support vector machine (SVM) as they allow for better subsequent interpretation without significant loss of accuracy.

Genomic risk prediction of complex human disease and its clinical application

Recent advances in genome-wide association studies have stimulated interest in the genomic prediction of disease risk, potentially enabling individual-level risk estimates for early intervention and improved diagnostic procedures. Here, we review recent findings and approaches to genomic prediction model construction and performance, then contrast the potential benefits of such models in two complex human diseases, aiding diagnosis in celiac disease and prospective risk prediction for cardiovascular disease. Early indications are that optimal application of genomic risk scores will differ substantially for each disease depending on underlying genetic architecture as well as current clinical and public health practice. As costs decline, genomic profiles become common, and popular understanding of risk and its communication improves, genomic risk will become increasingly useful for the individual and the clinician.

Performance of genetic risk factors in prediction of trichloroethylene induced hypersensitivity syndrome

Trichloroethylene induced hypersensitivity syndrome is dose-independent and potentially life threatening disease, which has become one of the serious occupational health issues and requires intensive treatment. To discover the genetic risk factors and evaluate the performance of risk prediction model for the disease, we conducted genomewide association study and replication study with total of 174 cases and 1761 trichloroethylene-tolerant controls. Fifty seven SNPs that exceeded the threshold for genome-wide significance (P < 5 × 10(-8)) were screened to relate with the disease, among which two independent SNPs were identified, that is rs2857281 at MICA (odds ratio, 11.92; P meta = 1.33 × 10(-37)) and rs2523557 between HLA-B and MICA (odds ratio, 7.33; P meta = 8.79 × 10(-35)). The genetic risk score with these two SNPs explains at least 20.9% of the disease variance and up to 32.5-fold variation in inter-individual risk. Combining of two SNPs as predictors for the disease would have accuracy of 80.73%, the area under receiver operator characteristic curves (AUC) scores was 0.82 with sensitivity of 74% and specificity of 85%, which was considered to have excellent discrimination for the disease, and could be considered for translational application for screening employees before exposure.

Genomic prediction of celiac disease targeting HLA-positive individuals

Background

Genomic prediction aims to leverage genome-wide genetic data towards better disease diagnostics and risk scores. We have previously published a genomic risk score (GRS) for celiac disease (CD), a common and highly heritable autoimmune disease, which differentiates between CD cases and population-based controls at a clinically-relevant predictive level, improving upon other gene-based approaches. HLA risk haplotypes, particularly HLA-DQ2.5, are necessary but not sufficient for CD, with at least one HLA risk haplotype present in up to half of most Caucasian populations. Here, we assess a genomic prediction strategy that specifically targets this common genetic susceptibility subtype, utilizing a supervised learning procedure for CD that leverages known HLA-DQ2.5 risk.

Methods

Using L1/L2-regularized support-vector machines trained on large European case-control datasets, we constructed novel CD GRSs specific to individuals with HLA-DQ2.5 risk haplotypes (GRS-DQ2.5) and compared them with the predictive power of the existing CD GRS (GRS14) as well as two haplotype-based approaches, externally validating the results in a North American case-control study.

Results

Consistent with previous observations, both the existing GRS14 and the GRS-DQ2.5 had better predictive performance than the HLA haplotype approaches. GRS-DQ2.5 models, based on directly genotyped or imputed markers, achieved similar levels of predictive performance (AUC = 0.718-0.73), which were substantially higher than those obtained from the DQ2.5 zygosity alone (AUC = 0.558), the HLA risk haplotype method (AUC = 0.634), or the generic GRS14 (AUC = 0.679). In a screening model of at-risk individuals, the GRS-DQ2.5 lowered the number of unnecessary follow-up tests for CD across most sensitivity levels. Relative to a baseline implicating all DQ2.5-positive individuals for follow-up, the GRS-DQ2.5 resulted in a net saving of 2.2 unnecessary follow-up tests for each justified test while still capturing 90 % of DQ2.5-positive CD cases.

Conclusions

Genomic risk scores for CD that target genetically at-risk sub-groups improve predictive performance beyond traditional approaches and may represent a useful strategy for prioritizing individuals at increased risk of disease, thus potentially reducing unnecessary follow-up diagnostic tests.

Electronic supplementary material

The online version of this article (doi:10.1186/s13073-015-0196-5) contains supplementary material, which is available to authorized users.