Differential expression analysis for RNAseq using Poisson mixed models

Abstract

Identifying differentially expressed (DE) genes from RNA sequencing (RNAseq) studies is among the most common analyses in genomics. However, RNAseq DE analysis presents several statistical and computational challenges, including over-dispersed read counts and, in some settings, sample non-independence. Previous count-based methods rely on simple hierarchical Poisson models (e.g. negative binomial) to model independent over-dispersion, but do not account for sample non-independence due to relatedness, population structure and/or hidden confounders. Here, we present a Poisson mixed model with two random effects terms that account for both independent over-dispersion and sample non-independence. We also develop a scalable sampling-based inference algorithm using a latent variable representation of the Poisson distribution. With simulations, we show that our method properly controls for type I error and is generally more powerful than other widely used approaches, except in small samples (n <15) with other unfavorable properties (e.g. small effect sizes). We also apply our method to three real datasets that contain related individuals, population stratification or hidden confounders. Our results show that our method increases power in all three data compared to other approaches, though the power gain is smallest in the smallest sample (n = 6). Our method is implemented in MACAU, freely available at www.xzlab.org/software.html.

Covariance-insured screening

“…Despite this promising outlook, analytic methods remain insufficient for achieving truly personalized medicine. The standard procedure is to evaluate one gene at a time, which results in low statistical power to identify the disease-associated genes [28]. Thus, more accurate models that leverage the large amounts of genomic data now available are in great demand.…”

Statistical Analysis of Spatial Expression Pattern for Spatially Resolved Transcriptomic Studies

“…However, analyzing normalized expression data can be suboptimal as this approach fails to account for the mean-variance relationship existed in raw counts, leading to a potential loss of power 14 . Indeed, similar loss of power has been well documented for methods that can only analyze normalized data in many other omics sequencing studies 15,16 . Besides direct modeling of count data, identifying SE genes also requires the development of statistical methods that can produce well calibrated p-values to ensure proper control of type I error.…”

“…We then used the hierarchical agglomerative clustering algorithm in the R package amap (v0. [8][9][10][11][12][13][14][15][16][17] to cluster identified SE genes detected by SPARK into five groups. Afterwards, we summarized the gene expression patterns by using the expression level of the five cluster centers (Fig.…”

Subset-Based Analysis Using Gene-Environment Interactions for Discovery of Genetic Associations across Multiple Studies or Phenotypes

“…These properties make pASTA a powerful tool to identify candidate SNPs that are potentially associated with the trait(s) of interest in the screening step of a GWAS, which then can be followed by deeper characterization/interpretation of the associations through other methods, such as biological validation. For example, RNA sequencing experiments can be performed to examine whether the candidate SNPs affect the expression level of the target gene [34]. Crispr-Cas9 knockout screening can be carried out in human cell lines to determine the functional significance of these candidate SNPs [35].…”