Bugli, CélineVast, MadelineMadelineVast2025-05-142025-05-142025-05-142024https://hdl.handle.net/2078.2/36789RNA sequencing is a technology used in transcriptomics. It allows to survey the entire transcriptome without prior knowledge of the sequences. This technology is commonly used in experiments to detect genes that are differentially expressed between different conditions. However, the most popular R packages used to this end do not use models appropriate for experimental design with dependence between the observations. Several approaches have been proposed to perform differential expression analysis on RNA-Seq data with dependence between the observations. Most of them use either generalized linear mixed models or linear mixed models on transformed data. This thesis aims to compare the package available in R for the differential expression analysis of RNA-Seq data using linear mixed models, lmerSeq and variancePartition. This Master’s thesis has highlighted that the main difference between the two packages is their approaches to take into account the mean-variance relationship in RNA-Seq data. In addition, although lmerSeq performs better for estimating model parameters, the two packages perform similarly for contrast testing of experiments with more than 20 samples. However, the false discovery rate is above the target value when variancePartition is used for a sample size below 20.RNA-SeqLinear mixed modelstext::thesis::master thesisthesis:44166