Analysis of untyped SNPs: maximum likelihood and imputation methods

Genetic signature of differentiated thyroid carcinoma susceptibility: a machine learning approach

To identify a peculiar genetic combination predisposing to differentiated thyroid carcinoma (DTC), we selected a set of single nucleotide polymorphisms (SNPs) associated with DTC risk, considering polygenic risk score (PRS), Bayesian statistics and a machine learning (ML) classifier to describe cases and controls in three different datasets. Dataset 1 (649 DTC, 431 controls) has been previously genotyped in a genome-wide association study (GWAS) on Italian DTC. Dataset 2 (234 DTC, 101 controls) and dataset 3 (404 DTC, 392 controls) were genotyped. Associations of 171 SNPs reported to predispose to DTC in candidate studies were extracted from the GWAS of dataset 1, followed by replication of SNPs associated with DTC risk (P < 0.05) in dataset 2. The reliability of the identified SNPs was confirmed by PRS and Bayesian statistics after merging the three datasets. SNPs were used to describe the case/control state of individuals by ML classifier. Starting from 171 SNPs associated with DTC, 15 were positive in both datasets 1 and 2. Using these markers, PRS revealed that individuals in the fifth quintile had a seven-fold increased risk of DTC than those in the first. Bayesian inference confirmed that the selected 15 SNPs differentiate cases from controls. Results were corroborated by ML, finding a maximum AUC of about 0.7. A restricted selection of only 15 DTC-associated SNPs is able to describe the inner genetic structure of Italian individuals, and ML allows a fair prediction of case or control status based solely on the individual genetic background.

Integrating variant functional annotation scores have varied abilities to improve power of genome-wide association studies

Functional annotations have the potential to increase power of genome-wide association studies (GWAS) by prioritizing variants according to their biological function, but this potential has not been well studied. We comprehensively evaluated all 1132 traits in the UK Biobank whose SNP-heritability estimates were given “medium” or “high” labels by Neale’s lab. For each trait, we integrated GWAS summary statistics of close to 8 million common variants (minor allele frequency \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$>1\%$$\end{document}>1%) with either their 75 individual functional scores or their meta-scores, using three different data-integration methods. Overall, the number of new genome-wide significant findings after data-integration increases as a trait SNP-heritability estimate increases. However, there is a trade-off between new findings and loss of baseline GWAS findings, resulting in similar total numbers of significant findings between using GWAS alone and integrating GWAS with functional scores, across all 1132 traits analyzed and all three data-integration methods considered. Our findings suggest that, even with the current biobank-level sample size, more informative functional scores and/or new data-integration methods are needed to further improve the power of GWAS of common variants. For example, studying variants in coding sequence and obtaining cell-type-specific scores are potential future directions.

A new approach to testing mediation of the microbiome at both the community and individual taxon levels

MOTIVATION

Understanding whether and which microbes played a mediating role between an exposure and a disease outcome are essential for researchers to develop clinical interventions to treat the disease by modulating the microbes. Existing methods for mediation analysis of the microbiome are often limited to a global test of community-level mediation or selection of mediating microbes without control of the false discovery rate (FDR). Further, while the null hypothesis of no mediation at each microbe is a composite null that consists of three types of null, most existing methods treat the microbes as if they were all under the same type of null, leading to excessive false positive results.

RESULTS

We propose a new approach based on inverse regression that regresses the microbiome data at each taxon on the exposure and the exposure-adjusted outcome. Then, the P-values for testing the coefficients are used to test mediation at both the community and individual taxon levels. This approach fits nicely into our Linear Decomposition Model (LDM) framework, so our new method LDM-med, implemented in the LDM framework, enjoys all the features of the LDM, e.g. allowing an arbitrary number of taxa to be tested simultaneously, supporting continuous, discrete, or multivariate exposures and outcomes (including survival outcomes), and so on. Using extensive simulations, we showed that LDM-med always preserved the FDR of testing individual taxa and had adequate sensitivity; LDM-med always controlled the type I error of the global test and had compelling power over existing methods. The flexibility of LDM-med for a variety of mediation analyses is illustrated by an application to a murine microbiome dataset, which identified several plausible mediating taxa.

AVAILABILITY AND IMPLEMENTATION

Our new method has been added to our R package LDM, which is available on GitHub at https://github.com/yijuanhu/LDM.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A likelihood-based approach to transcriptome association analysis

Elucidating the mechanistic underpinnings of genetic associations with complex traits requires formally characterizing and testing associated cell and tissue-specific expression profiles. New opportunities exist to bolster this investigation with the growing numbers of large publicly available omics level data resources. Herein, we describe a fully likelihood-based strategy to leveraging external resources in the setting that expression profiles are partially or fully unobserved in a genetic association study. A general framework is presented to accommodate multiple data types, and strategies for implementation using existing software packages are described. The method is applied to an investigation of the genetics of evoked inflammatory response in cardiovascular disease research. Simulation studies suggest appropriate type-1 error control and power gains compared to single regression imputation, the most commonly applied practice in this setting.

Multi-SNP Haplotype Analysis Methods for Association Analysis

Haplotype analysis forms the basis of much of genetic association analysis using both related and unrelated individuals (we concentrate on unrelated). For example, haplotype analysis indirectly underlies the SNP imputation methods that are used for testing trait associations with known but unmeasured variants and for performing collaborative post-GWAS meta-analysis. This chapter is focused on the direct use of haplotypes in association testing. It reviews the rationale for haplotype-based association testing, discusses statistical issues related to haplotype uncertainty that affect the analysis, then gives practical guidance for testing haplotype-based associations with phenotype or outcome trait, first of candidate gene regions and then for the genome as a whole. Haplotypes are interesting for two reasons, first they may be in closer LD with a causal variant than any single measured SNP, and therefore may enhance the coverage value of the genotypes over single SNP analysis. Second, haplotypes may themselves be the causal variants of interest and some solid examples of this have appeared in the literature.This chapter discusses three possible approaches to incorporation of SNP haplotype analysis into generalized linear regression models: (1) a simple substitution method involving imputed haplotypes, (2) simultaneous maximum likelihood (ML) estimation of all parameters, including haplotype frequencies and regression parameters, and (3) a simplified approximation to full ML for case-control data.Examples of the various approaches for a haplotype analysis of a candidate gene are provided. We compare the behavior of the approximation-based methods and argue that in most instances the simpler methods hold up well in practice. We also describe the practical implementation of haplotype risk estimation genome-wide and discuss several shortcuts that can be used to speed up otherwise potentially very intensive computational requirements.

Assessing and comparison of different machine learning methods in parent-offspring trios for genotype imputation

Genotype imputation is an important tool for prediction of unknown genotypes for both unrelated individuals and parent-offspring trios. Several imputation methods are available and can either employ universal machine learning methods, or deploy algorithms dedicated to infer missing genotypes. In this research the performance of eight machine learning methods: Support Vector Machine, K-Nearest Neighbors, Extreme Learning Machine, Radial Basis Function, Random Forest, AdaBoost, LogitBoost, and TotalBoost compared in terms of the imputation accuracy, computation time and the factors affecting imputation accuracy. The methods employed using real and simulated datasets to impute the un-typed SNPs in parent-offspring trios. The tested methods show that imputation of parent-offspring trios can be accurate. The Random Forest and Support Vector Machine were more accurate than the other machine learning methods. The TotalBoost performed slightly worse than the other methods.The running times were different between methods. The ELM was always most fast algorithm. In case of increasing the sample size, the RBF requires long imputation time.The tested methods in this research can be an alternative for imputation of un-typed SNPs in low missing rate of data. However, it is recommended that other machine learning methods to be used for imputation.

Comparison of three boosting methods in parent-offspring trios for genotype imputation using simulation study

Background

Genotype imputation is an important process of predicting unknown genotypes, which uses reference population with dense genotypes to predict missing genotypes for both human and animal genetic variations at a low cost. Machine learning methods specially boosting methods have been used in genetic studies to explore the underlying genetic profile of disease and build models capable of predicting missing values of a marker.

Methods

In this study strategies and factors affecting the imputation accuracy of parent-offspring trios compared from lower-density SNP panels (5 K) to high density (10 K) SNP panel using three different Boosting methods namely TotalBoost (TB), LogitBoost (LB) and AdaBoost (AB). The methods employed using simulated data to impute the un-typed SNPs in parent-offspring trios. Four different datasets of G1 (100 trios with 5 k SNPs), G2 (100 trios with 10 k SNPs), G3 (500 trios with 5 k SNPs), and G4 (500 trio with 10 k SNPs) were simulated. In four datasets all parents were genotyped completely, and offspring genotyped with a lower density panel.

Results

Comparison of the three methods for imputation showed that the LB outperformed AB and TB for imputation accuracy. The time of computation were different between methods. The AB was the fastest algorithm. The higher SNP densities resulted the increase of the accuracy of imputation. Larger trios (i.e. 500) was better for performance of LB and TB.

Conclusions

The conclusion is that the three methods do well in terms of imputation accuracy also the dense chip is recommended for imputation of parent-offspring trios.

A Likelihood-Based Framework for Association Analysis of Allele-Specific Copy Numbers

Copy number variants (CNVs) and single nucleotide polymorphisms (SNPs) co-exist throughout the human genome and jointly contribute to phenotypic variations. Thus, it is desirable to consider both types of variants, as characterized by allele-specific copy numbers (ASCNs), in association studies of complex human diseases. Current SNP genotyping technologies capture the CNV and SNP information simultaneously via fluorescent intensity measurements. The common practice of calling ASCNs from the intensity measurements and then using the ASCN calls in downstream association analysis has important limitations. First, the association tests are prone to false-positive findings when differential measurement errors between cases and controls arise from differences in DNA quality or handling. Second, the uncertainties in the ASCN calls are ignored. We present a general framework for the integrated analysis of CNVs and SNPs, including the analysis of total copy numbers as a special case. Our approach combines the ASCN calling and the association analysis into a single step while allowing for differential measurement errors. We construct likelihood functions that properly account for case-control sampling and measurement errors. We establish the asymptotic properties of the maximum likelihood estimators and develop EM algorithms to implement the corresponding inference procedures. The advantages of the proposed methods over the existing ones are demonstrated through realistic simulation studies and an application to a genome-wide association study of schizophrenia. Extensions to next-generation sequencing data are discussed.

Integrative analysis of sequencing and array genotype data for discovering disease associations with rare mutations

In the large cohorts that have been used for genome-wide association studies (GWAS), it is prohibitively expensive to sequence all cohort members. A cost-effective strategy is to sequence subjects with extreme values of quantitative traits or those with specific diseases. By imputing the sequencing data from the GWAS data for the cohort members who are not selected for sequencing, one can dramatically increase the number of subjects with information on rare variants. However, ignoring the uncertainties of imputed rare variants in downstream association analysis will inflate the type I error when sequenced subjects are not a random subset of the GWAS subjects. In this article, we provide a valid and efficient approach to combining observed and imputed data on rare variants. We consider commonly used gene-level association tests, all of which are constructed from the score statistic for assessing the effects of individual variants on the trait of interest. We show that the score statistic based on the observed genotypes for sequenced subjects and the imputed genotypes for nonsequenced subjects is unbiased. We derive a robust variance estimator that reflects the true variability of the score statistic regardless of the sampling scheme and imputation quality, such that the corresponding association tests always have correct type I error. We demonstrate through extensive simulation studies that the proposed tests are substantially more powerful than the use of accurately imputed variants only and the use of sequencing data alone. We provide an application to the Women's Health Initiative. The relevant software is freely available.

A mega-analysis of genome-wide association studies for major depressive disorder

Prior genome-wide association studies (GWAS) of major depressive disorder (MDD) have met with limited success. We sought to increase statistical power to detect disease loci by conducting a GWAS mega-analysis for MDD. In the MDD discovery phase, we analyzed more than 1.2 million autosomal and X chromosome single-nucleotide polymorphisms (SNPs) in 18 759 independent and unrelated subjects of recent European ancestry (9240 MDD cases and 9519 controls). In the MDD replication phase, we evaluated 554 SNPs in independent samples (6783 MDD cases and 50 695 controls). We also conducted a cross-disorder meta-analysis using 819 autosomal SNPs with P<0.0001 for either MDD or the Psychiatric GWAS Consortium bipolar disorder (BIP) mega-analysis (9238 MDD cases/8039 controls and 6998 BIP cases/7775 controls). No SNPs achieved genome-wide significance in the MDD discovery phase, the MDD replication phase or in pre-planned secondary analyses (by sex, recurrent MDD, recurrent early-onset MDD, age of onset, pre-pubertal onset MDD or typical-like MDD from a latent class analyses of the MDD criteria). In the MDD-bipolar cross-disorder analysis, 15 SNPs exceeded genome-wide significance (P<5 × 10(-8)), and all were in a 248 kb interval of high LD on 3p21.1 (chr3:52 425 083-53 822 102, minimum P=5.9 × 10(-9) at rs2535629). Although this is the largest genome-wide analysis of MDD yet conducted, its high prevalence means that the sample is still underpowered to detect genetic effects typical for complex traits. Therefore, we were unable to identify robust and replicable findings. We discuss what this means for genetic research for MDD. The 3p21.1 MDD-BIP finding should be interpreted with caution as the most significant SNP did not replicate in MDD samples, and genotyping in independent samples will be needed to resolve its status.

Optimal methods for using posterior probabilities in association testing

OBJECTIVE

The use of haplotypes to impute the genotypes of unmeasured single nucleotide variants continues to rise in popularity. Simulation results suggest that the use of the dosage as a one-dimensional summary statistic of imputation posterior probabilities may be optimal both in terms of statistical power and computational efficiency; however, little theoretical understanding is available to explain and unify these simulation results. In our analysis, we provide a theoretical foundation for the use of the dosage as a one-dimensional summary statistic of genotype posterior probabilities from any technology.

METHODS

We analytically evaluate the dosage, mode and the more general set of all one-dimensional summary statistics of two-dimensional (three posterior probabilities that must sum to 1) genotype posterior probability vectors.

RESULTS

We prove that the dosage is an optimal one-dimensional summary statistic under a typical linear disease model and is robust to violations of this model. Simulation results confirm our theoretical findings.

CONCLUSIONS

Our analysis provides a strong theoretical basis for the use of the dosage as a one-dimensional summary statistic of genotype posterior probability vectors in related tests of genetic association across a wide variety of genetic disease models.

A genome-wide scan for breast cancer risk haplotypes among African American women

Genome-wide association studies (GWAS) simultaneously investigating hundreds of thousands of single nucleotide polymorphisms (SNP) have become a powerful tool in the investigation of new disease susceptibility loci. Haplotypes are sometimes thought to be superior to SNPs and are promising in genetic association analyses. The application of genome-wide haplotype analysis, however, is hindered by the complexity of haplotypes themselves and sophistication in computation. We systematically analyzed the haplotype effects for breast cancer risk among 5,761 African American women (3,016 cases and 2,745 controls) using a sliding window approach on the genome-wide scale. Three regions on chromosomes 1, 4 and 18 exhibited moderate haplotype effects. Furthermore, among 21 breast cancer susceptibility loci previously established in European populations, 10p15 and 14q24 are likely to harbor novel haplotype effects. We also proposed a heuristic of determining the significance level and the effective number of independent tests by the permutation analysis on chromosome 22 data. It suggests that the effective number was approximately half of the total (7,794 out of 15,645), thus the half number could serve as a quick reference to evaluating genome-wide significance if a similar sliding window approach of haplotype analysis is adopted in similar populations using similar genotype density.

Multi-SNP haplotype analysis methods for association analysis

This chapter reviews the rationale for the use of haplotypes in association-based testing, discusses statistical issues related to haplotype uncertainty that complicate the analysis, then gives practical guidance for testing haplotype-based associations with phenotype or outcome trait, first of candidate gene regions and then for the genome as a whole. Haplotypes are interesting for two reasons: First, they may be in closer LD with a causal variant than any single measured SNP, and therefore may enhance the coverage value of the genotypes over single SNP analysis. Second, haplotypes may themselves be the causal variants of interest and some solid examples of this have appeared in the literature. This chapter discusses three possible approaches to incorporation of SNP haplotype analysis into generalized linear regression models: (1) a simple substitution method involving imputed haplotypes; (2) simultaneous maximum likelihood (ML) estimation of all parameters, including haplotype frequencies and regression parameters; and (3) a simplified approximation to full ML for case-control data. Examples of the various approaches for a haplotype analysis of a candidate gene are provided. We compare the behavior of the approximation-based methods and show that in most instances the simpler methods hold up well in practice. We also describe the practical implementation of genome-wide haplotype risk estimation and discuss several shortcuts that can be used to speed up otherwise potentially very intensive computational requirements.