RnaSeqSampleSize

RnaSeqSampleSize: Negative Binomial-Based Sample Size Estimation for RNA-Seq

RnaSeqSampleSize estimates sample size and statistical power for differential gene expression analysis in RNA-Seq experiments using negative binomial models parameterized by gene-specific read counts and dispersion estimates.

Key Features:

• Negative Binomial Model-Based Estimation: Applies a negative binomial distribution to model RNA-Seq count data, accounting for variability and overdispersion in gene expression measurements.
• Multiple Testing Control: Incorporates false discovery rate (FDR) control to support power calculations across thousands of simultaneously tested genes.
• Distribution-Based Estimation: Derives gene-specific read count and dispersion parameters from empirical RNA-Seq datasets, including The Cancer Genome Atlas (TCGA).
• Gene and Pathway-Specific Analysis: Enables targeted sample size and power estimation for selected genes or biological pathways.
• Parameter Optimization: Optimizes model parameters to improve precision and reliability of power and sample size estimates.

Scientific Applications:

• Differential Expression Study Design: Determines sample sizes required to achieve specified statistical power for detecting differential gene expression in RNA-Seq experiments while controlling FDR.

Methodology:

Gene-specific read counts and dispersion parameters are estimated from reference RNA-Seq datasets such as TCGA. These parameters are modeled using a negative binomial distribution to compute statistical power and required sample sizes under multiple hypothesis testing with FDR control.