,
Yubai Yuan [ctb],
Annie Qu [ctb]| Title: | Differentially Expressed Heterogeneous Overdispersion Gene Test for Count Data |
|---|---|
| Description: | Implements a generalized linear model approach for detecting differentially expressed genes across treatment groups in count data. The package supports both quasi-Poisson and negative binomial models to handle overdispersion, ensuring robust identification of differential expression. It allows for the inclusion of treatment effects and gene-wise covariates, as well as normalization factors for accurate scaling across samples. Additionally, it incorporates statistical significance testing with options for p-value adjustment and log2 fold range thresholds, making it suitable for RNA-seq analysis. |
| Authors: | Qi Xu [aut],
Arlina Shen [cre] ,
Yubai Yuan [ctb],
Annie Qu [ctb] |
| Maintainer: | Arlina Shen <[email protected]> |
| License: | GPL-3 |
| Version: | 0.99.0 |
| Built: | 2024-09-14 07:33:20 UTC |
| Source: | https://github.com/ahshen26/dehogt |
Differentially Expressed Heterogeneous Overdispersion Genes Testing for Count Data This script implements the main function of the proposed method in the above paper
dehogt_func( data, treatment, norm_factors = NULL, covariates = NULL, dist = "qpois", padj = TRUE, pval_thre = 0.05, l2fc = FALSE, l2fc_thre = 1, num_cores = 1 )dehogt_func( data, treatment, norm_factors = NULL, covariates = NULL, dist = "qpois", padj = TRUE, pval_thre = 0.05, l2fc = FALSE, l2fc_thre = 1, num_cores = 1 )
data |
A matrix of gene expression data where rows represent genes and columns represent samples. |
treatment |
A vector specifying the treatment conditions for each sample. |
norm_factors |
An optional vector of normalization factors for each sample. Default is NULL, which assumes equal normalization factors. |
covariates |
An optional matrix of gene-wise covariates. Default is NULL. |
dist |
The distribution family for the GLM. Can be "qpois" for quasi-Poisson or "negbin" for negative binomial. Default is "qpois". |
padj |
Logical value indicating whether to adjust p-values using the Benjamini-Hochberg (BH) procedure. Default is TRUE. |
pval_thre |
The threshold for identifying differentially expressed genes based on adjusted p-values. Default is 0.05. |
l2fc |
Logical value indicating whether to consider log2 fold change for identifying differentially expressed genes. Default is FALSE. |
l2fc_thre |
The threshold for log2 fold change in identifying differentially expressed genes. Default is 1. |
num_cores |
The number of CPU cores to use for parallel computing. Default is 1. |
A list containing:
DE_idx |
A logical vector indicating differentially expressed genes. |
pvals |
A numeric vector of p-values for each gene. |
log2fc |
A numeric vector of log2 fold changes for each gene. |
# simulate gene expression data data <- matrix(rpois(1000, 10), nrow = 100, ncol = 10) # simulate random treatment assignments treatment <- sample(0:1, 10, replace = TRUE) # Run main function with parallel computing using 2 cores result <- dehogt_func(data, treatment, num_cores = 2)# simulate gene expression data data <- matrix(rpois(1000, 10), nrow = 100, ncol = 10) # simulate random treatment assignments treatment <- sample(0:1, 10, replace = TRUE) # Run main function with parallel computing using 2 cores result <- dehogt_func(data, treatment, num_cores = 2)