Rare variants analysis using penalization methods for whole genome sequence data

Abstract

BackgroundAvailability of affordable and accessible whole genome sequencing for biomedical applications poses a number of statistical challenges and opportunities, particularly related to the analysis of rare variants and sparseness of the data. Although efforts have been devoted to address these challenges, the performance of statistical methods for rare variants analysis still needs further consideration.ResultWe introduce a new approach that applies restricted principal component analysis with convex penalization and then selects the best predictors of a phenotype by a concave penalized regression model, while estimating the impact of each genomic region on the phenotype. Using simulated data, we show that the proposed method maintains good power for association testing while keeping the false discovery rate low under a verity of genetic architectures. Illustrative data analyses reveal encouraging result of this method in comparison with other commonly applied methods for rare variants analysis.ConclusionBy taking into account linkage disequilibrium and sparseness of the data, the proposed method improves power and controls the false discovery rate compared to other commonly applied methods for rare variant analyses.

Identification of Rare Variants in Metabolites of the Carnitine Pathway by Whole Genome Sequencing Analysis

“…To address the multitude of analytic challenges confronted when analyzing rare variants across the genome, we applied a new statistical approach based on penalization methods by Yazdani et al. []. This approach follows a two‐stage analysis.…”

“…First, there are a very large number of rare variants and many of these variants are limited to a single individual. Therefore, usual single site methods comparing mean values or frequencies do not have sufficient statistical power and are often not appropriate or feasible [Lee et al., ; Yazdani et al., ]. Second, outside of genes and a few well‐characterized regulatory elements, the role of much of the genome in normal development, metabolism, and physiology remains unknown.…”

“…We use whole genome sequence data and rare variant analysis methods introduced by Yazdani et al. [] called Convex‐Concave Rare variant Selection (CCRS) to investigate a model subset of the human serum metabolome, specifically the carnitine‐related metabolites. Carnitine is synthesized in the liver and kidney and is required for the transport of fatty acids within mitochondria for energy production [Steiber et al., ].…”

On the association analysis of genome‐sequencing data: A spatial clustering approach for partitioning the entire genome into nonoverlapping windows

“…Most of the current methodology on grouping algorithms for whole genome data has focused on window‐based approaches (e.g., Morrison et al., , Panoutsopoulou, Tachmazidou, & Zeggini, , Yazdani, Yazdani, & Boerwinkle, ). The intuition behind window‐based approaches is similar to gene‐based tests, namely that variants that lie within defined windows are jointly tested for association.…”

An improved burden-test pipeline for identifying associations from rare germline and somatic variants

“…Some approaches weight the given variants before statistical tests, such as the rare variant weighted aggregated statistic ( RWAS ) [ 7 ], the likelihood ratio test ( LRT ) [ 8 ] and two weighted score tests with branching under ratios ( BUR ) and likelihood-based model branching ( LiMB ), respectively [ 9 ]. Regression is another idea to refine the given variants, such as the kernel-based association test ( KBAT ) [ 10 ], the sequence kernel association tests ( SKAT ) [ 11 , 12 ] and convex-concave rare variant selection method ( CCRS ) [ 13 ]. RareCover is considered the first algorithmic selection approach, which filters the variants via a χ 2 aggregation greedy strategy [ 14 ].…”