Power analysis and sample size estimation for sequence-based association studies

Study design and the sampling of deleterious rare variants in biobank-scale datasets

One key component of study design in population genetics is the "geographic breadth" of a sample (i.e., how broad a region across which individuals are sampled). How the geographic breadth of a sample impacts observations of rare, deleterious variants is unclear, even though such variants are of particular interest for biomedical and evolutionary applications. Here, in order to gain insight into the effects of sample design on ascertained genetic variants, we formulate a stochastic model of dispersal, genetic drift, selection, mutation, and geographically concentrated sampling. We use this model to understand the effects of the geographic breadth of sampling effort on the discovery of negatively selected variants. We find that samples which are more geographically broad will discover a greater number variants as compared geographically narrow samples (an effect we label "discovery"); though the variants will be detected at lower average frequency than in narrow samples (e.g. as singletons, an effect we label "dilution"). Importantly, these effects are amplified for larger sample sizes and moderated by the magnitude of fitness effects. We validate these results using both population genetic simulations and empirical analyses in the UK Biobank. Our results are particularly important in two contexts: the association of large-effect rare variants with particular phenotypes and the inference of negative selection from allele frequency data. Overall, our findings emphasize the importance of considering geographic breadth when designing and carrying out genetic studies, especially at biobank scale.

Ravages: An R package for the simulation and analysis of rare variants in multicategory phenotypes

Current software packages for the analysis and the simulations of rare variants are only available for binary and continuous traits. Ravages provides solutions in a single R package to perform rare variant association tests for multicategory, binary and continuous phenotypes, to simulate datasets under different scenarios and to compute statistical power. Association tests can be run in the whole genome thanks to C++ implementation of most of the functions, using either RAVA-FIRST, a recently developed strategy to filter and analyse genome-wide rare variants, or user-defined candidate regions. Ravages also includes a simulation module that generates genetic data for cases who can be stratified into several subgroups and for controls. Through comparisons with existing programmes, we show that Ravages complements existing tools and will be useful to study the genetic architecture of complex diseases. Ravages is available on the CRAN at https://cran.r-project.org/web/packages/Ravages/ and maintained on Github at https://github.com/genostats/Ravages.

Detecting Clustered Independent Rare Variant Associations Using Genetic Algorithms

The availability of an increasing collection of sequencing data provides the opportunity to study genetic variation with an unprecedented level of detail. There is much interest in uncovering the role of rare variants and their contribution to disease. However, detecting associations of rare variants with small minor allele frequencies (MAF) and modest effects remains a challenge for rare variant association methods. Due to this low signal-to-noise ratio, most methods are underpowered to detect associations even when conducting rare variant association tests at the gene level. We present a new method for detecting rare variant associations. The algorithm consists of two steps. In the first step, a genetic algorithm searches for a promising genomic region containing a collection of genes with causal rare variants. In the second step, a genetic algorithm aims at removing false positives from the located genomic region. We tested the proposed method with a collection of datasets obtained from real exome data. The proposed method possesses sufficient power for detecting associations of rare variants with complex phenotypes. This method can be used for studying the contribution of rare variants with complex disease, particularly in cases where single-variant or gene-based tests are underpowered.

U-PASS: unified power analysis and forensics for qualitative traits in genetic association studies

Summary

Despite the availability of existing calculators for statistical power analysis in genetic association studies, there has not been a model-invariant and test-independent tool that allows for both planning of prospective studies and systematic review of reported findings. In this work, we develop a web-based application U-PASS (Unified Power analysis of ASsociation Studies), implementing a unified framework for the analysis of common association tests for binary qualitative traits. The application quantifies the shared asymptotic power limits of the common association tests, and visualizes the fundamental statistical trade-off between risk allele frequency and odds ratio. The application also addresses the applicability of asymptotics-based power calculations in finite samples, and provides guidelines for single-SNP-based association tests. In addition to designing prospective studies, U-PASS enables researchers to retrospectively assess the statistical validity of previously reported associations.

Availability and implementation

U-PASS is an open-source R Shiny application. A live instance is hosted at https://power.stat.lsa.umich.edu. Source is available on https://github.com/Pill-GZ/U-PASS.

Supplementary information

Supplementary data are available at Bioinformatics online.

Power and sample size calculations for high-throughput sequencing-based experiments

Power/sample size (power) analysis estimates the likelihood of successfully finding the statistical significance in a data set. There has been a growing recognition of the importance of power analysis in the proper design of experiments. Power analysis is complex, yet necessary for the success of large studies. It is important to design a study that produces statistically accurate and reliable results. Power computation methods have been well established for both microarray-based gene expression studies and genotyping microarray-based genome-wide association studies. High-throughput sequencing (HTS) has greatly enhanced our ability to conduct biomedical studies at the highest possible resolution (per nucleotide). However, the complexity of power computations is much greater for sequencing data than for the simpler genotyping array data. Research on methods of power computations for HTS-based studies has been recently conducted but is not yet well known or widely used. In this article, we describe the power computation methods that are currently available for a range of HTS-based studies, including DNA sequencing, RNA-sequencing, microbiome sequencing and chromatin immunoprecipitation sequencing. Most importantly, we review the methods of power analysis for several types of sequencing data and guide the reader to the relevant methods for each data type.

SimPEL: Simulation-based power estimation for sequencing studies of low-prevalence conditions

Power estimations are important for optimizing genotype-phenotype association study designs. However, existing frameworks are designed for common disorders, and thus ill-suited for the inherent challenges of studies for low-prevalence conditions such as rare diseases and infrequent adverse drug reactions. These challenges include small sample sizes and the need to leverage genetic annotation resources in association analyses for the purpose of ranking potential causal genes. We present SimPEL, a simulation-based program providing power estimations for the design of low-prevalence condition studies. SimPEL integrates the usage of gene annotation resources for association analyses. Customizable parameters, including the penetrance of the putative causal allele and the employed pathogenic scoring system, allow SimPEL to realistically model a large range of study designs. To demonstrate the effects of various parameters on power, we estimated the power of several simulated designs using SimPEL and captured power trends in agreement with observations from current literature on low-frequency condition studies. SimPEL, as a tool, provides researchers studying low-frequency conditions with an intuitive and highly flexible avenue for statistical power estimation. The platform-independent "batteries included" executable and default input files are available at https://github.com/precisionomics/SimPEL.

Power Analysis for Genetic Association Test (PAGEANT) provides insights to challenges for rare variant association studies

Motivation

Genome-wide association studies are now shifting focus from analysis of common to rare variants. As power for association testing for individual rare variants may often be low, various aggregate level association tests have been proposed to detect genetic loci. Typically, power calculations for such tests require specification of large number of parameters, including effect sizes and allele frequencies of individual variants, making them difficult to use in practice. We propose to approximate power to a varying degree of accuracy using a smaller number of key parameters, including the total genetic variance explained by multiple variants within a locus.

Results

We perform extensive simulation studies to assess the accuracy of the proposed approximations in realistic settings. Using these simplified power calculations, we develop an analytic framework to obtain bounds on genetic architecture of an underlying trait given results from genome-wide association studies with rare variants. Finally, we provide insights into the required quality of annotation/functional information for identification of likely causal variants to make meaningful improvement in power.

Availability and implementation

A shiny application that allows a variety of Power Analysis of GEnetic AssociatioN Tests (PAGEANT), in R is made publicly available at https://andrewhaoyu.shinyapps.io/PAGEANT/.

Contact

nilanjan@jhu.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

Postmortem Genetic Testing for Cardiac Ion Channelopathies in Stillbirths

BACKGROUND

Although stillbirth is a significant health problem worldwide, the definitive cause of death remains elusive in many cases, despite detailed autopsy. In this study of partly explained and unexplained stillbirths, we used next-generation sequencing to examine an extended panel of 35 candidate genes known to be associated with ion channel disorders and sudden cardiac death.

METHODS AND RESULTS

We examined tissue from 242 stillbirths (≥22 weeks), including those where no definite cause of death could be confirmed after a full autopsy. We obtained high-quality DNA from 70 cases, which were then sequenced for a custom panel of 35 genes, 12 for inherited long- and short-QT syndrome genes (LQT1-LQT12 and SQT1-3), and 23 additional candidate genes derived from genome-wide association studies. We examined the functional significance of a selected variant by patch-clamp electrophysiological recording. No predicted damaging variants were identified in KCNQ1 (LQT1) or KCNH2 (LQT2). A rare putative pathogenic variant was found in KCNJ2(LQT7) in 1 case, and several novel variants of uncertain significance were observed. The KCNJ2 variant (p. R40Q), when assessed by whole-cell patch clamp, affected the function of the channel. There was no significant evidence of enrichment of rare predicted damaging variants within any of the candidate genes.

CONCLUSIONS

Although a causative link is unclear, 1 putative pathogenic and variants of uncertain significance variant resulting in cardiac channelopathies was identified in some cases of otherwise unexplained stillbirth, and these variants may have a role in fetal demise.

CLINICAL TRIAL REGISTRATION

URL: https://www.clinicaltrials.gov. Unique identifier: NCT01120886.

Comparison of protein-based cell-of-origin classification to the Lymph2Cx RNA assay in a cohort of diffuse large B-cell lymphomas in Malaysia

Aims

The cell of origin (COO) based molecular characterisation into germinal centre B-cell-like (GCB) and activated B-cell-like (ABC) subtypes are central to the pathogenesis and clinical course in diffuse large B-cell lymphoma (DLBCL). Globally, clinical laboratories employ pragmatic but less than ideal immunohistochemical (IHC) assay for COO classification. Novel RNA-based platforms using routine pathology samples are emerging as new gold standard and offer unique opportunities for assay standardisation for laboratories across the world. We evaluated our IHC protocols against RNA-based technologies to determine concordance; additionally, we gauged the impact of preanalytical variation on the performance of Lymph2Cx assay.

Methods

Diagnostic biopsies (n=104) were examined for COO classification, employing automated RNA digital quantification assay (Lymph2Cx). Results were equated against IHC-based COO categorisation. Assay performance was assessed through its impact on overall survival (OS).

Results

96 (92%) informative samples were labelled as GCB (38/96; 40%) and non-GCB (58/96; 60%) by IHC evaluation. Lymph2Cx catalogued 36/96 (37%) samples as GCB, 45/96 (47%) as ABC and 15/96 (16%) as unclassified. Lymph2Cx being reference, IHC protocol revealed sensitivity of 81% for ABC and 75% for GCB categorisation and positive predictive value of 81% versus 82%, respectively. Lymph2Cx-based COO classification performed superior to Hans algorithm in predicting OS (log rank test, p=0.017 vs p=0.212).

Conclusions

Our report show that current IHC-based protocols for COO classification of DLBCL at UKM Malaysia are in line with previously reported results and marked variation in preanalytical factors do not critically impact Lymph2Cx assay quality.

Guidelines for Large-Scale Sequence-Based Complex Trait Association Studies: Lessons Learned from the NHLBI Exome Sequencing Project

Massively parallel whole-genome sequencing (WGS) data have ushered in a new era in human genetics. These data are now being used to understand the role of rare variants in complex traits and to advance the goals of precision medicine. The technological and computing advances that have enabled us to generate WGS data on thousands of individuals have also outpaced our ability to perform analyses in scientifically and statistically rigorous and thoughtful ways. The past several years have witnessed the application of whole-exome sequencing (WES) to complex traits and diseases. From our analysis of NHLBI Exome Sequencing Project (ESP) data, not only have a number of important disease and complex trait association findings emerged, but our collective experience offers some valuable lessons for WGS initiatives. These include caveats associated with generating automated pipelines for quality control and analysis of rare variants; the importance of studying minority populations; sample size requirements and efficient study designs for identifying rare-variant associations; and the significance of incidental findings in population-based genetic research. With the ESP as an example, we offer guidance and a framework on how to conduct a large-scale association study in the era of WGS.

On Sample Size and Power Calculation for Variant Set-Based Association Tests

Sample size and power calculations are an important part of designing new sequence-based association studies. The recently developed SEQPower and SPS programs adopted computationally intensive Monte Carlo simulations to empirically estimate power for a series of variant set association (VSA) test methods including the sequence kernel association test (SKAT). It is desirable to develop methods that can quickly and accurately compute power without intensive Monte Carlo simulations. We will show that the computed power for SKAT based on the existing analytical approach could be inflated especially for small significance levels, which are often of primary interest for large-scale whole genome and exome sequencing projects. We propose a new χ(2) -approximation-based approach to accurately and efficiently compute sample size and power. In addition, we propose and implement a more accurate "exact" method to compute power, which is more efficient than the Monte Carlo approach though generally involves more computations than the χ(2) approximation method. The exact approach could produce very accurate results and be used to verify alternative approximation approaches. We implement the proposed methods in publicly available R programs that can be readily adapted when planning sequencing projects.

Genetic data simulators and their applications: an overview

Computer simulations have played an indispensable role in the development and applications of statistical models and methods for genetic studies across multiple disciplines. The need to simulate complex evolutionary scenarios and pseudo-datasets for various studies has fueled the development of dozens of computer programs with varying reliability, performance, and application areas. To help researchers compare and choose the most appropriate simulators for their studies, we have created the genetic simulation resources (GSR) website, which allows authors of simulation software to register their applications and describe them with more than 160 defined attributes. This article summarizes the properties of 93 simulators currently registered at GSR and provides an overview of the development and applications of genetic simulators. Unlike other review articles that address technical issues or compare simulators for particular application areas, we focus on software development, maintenance, and features of simulators, often from a historical perspective. Publications that cite these simulators are used to summarize both the applications of genetic simulations and the utilization of simulators.