Transcriptomics is one of the most exciting omics fields that enable researchers to study various diseases and biological systems. The abundance of transcriptome-related data has facilitated the design and implementation of hundreds of studies with considerable directions over the past years. Transcriptomics offers a deeper understanding of diseases and helps identify biomarkers in that disease. Two widely used data types in transcriptomics are microarray and RNA-Seq data. Most standard workflows for analyzing these data are often based on the R programming language. The R packages available for analyzing transcriptome data have enabled the creation of unique and innovative studies. This article will review the most important packages for analyzing microarray and RNA-Seq data.

It should be noted that the analysis of transcriptome data takes place in two vital phases:

• Data preprocessing
• Performing the actual analysis (such as identifying differentially expressed genes)

Data Preprocessing

Before initiating the analysis, it is crucial to preprocess the raw data to remove any noise and artifacts. For Microarray data, this preprocessing typically involves several key steps:

1. Background Correction: This step corrects for non-specific probe binding to enhance data accuracy.
2. Normalization: This process ensures that the data from different arrays are comparable by scaling them appropriately.
3. Quality Control: This involves assessing the data for technical issues or outliers that may affect the results.
4. Log Transformation: Log transformation is frequently used to stabilize variance and make identifying patterns easier.
5. Batch Effect Correction: When multiple batches of microarray experiments are conducted, it is essential to correct batch effects to avoid confounding results.

Several popular R packages for RNA-seq data analysis can also perform data preprocessing. We will discuss these packages in the following sections. RNA-seq data typically requires several preprocessing steps, including normalization, quality control, log transformation, and batch effect correction.

R Packages for Data Preprocessing

Since microarray raw data are mostly preprocessed with specific R packages, in this section, we will list important ones for this purpose:

• affy: This package is designed to analyze Affymetrix oligonucleotide arrays. It includes functions for background correction, normalization, and summarization of probe-level data.
• simpleaffy: An extension of the affy package, simpleaffy simplifies the preprocessing of Affymetrix data, offering functions for quality control, normalization, and summarization.
• oligo: This versatile package is capable of analyzing various high-throughput arrays, including oligonucleotide and SNP arrays. It provides essential tools for preprocessing, such as background correction and normalization.
• limma: The limma package, a widely trusted tool for analyzing gene expression data, offers robust functions for background correction, normalization, data exploration, and differential expression analysis.
• gcrma: This package implements the GC Robust Multi-array Average (GCRMA) method for background adjustment and normalization of Affymetrix microarray data.
• preprocessCore: This package includes functions for various normalization methods, including quantile normalization, widely used in microarray data preprocessing.
• beadarray: Designed for Illumina BeadArray data. The beadarray package provides tools for reading raw data, quality assessment, normalization, and summarization.
• ArrayTools: A suite of tools for preprocessing, quality control, and microarray data normalization.

Differential Expression Analysis

Differential Expression Analysis is one of the most popular analyses observed in research articles and performed on microarray and RNA-seq data. The goal of this analysis is to find genes that are dysregulated (either up- or Down-regulated) in a condition like cancer compared to normal. Of course, it can be performed on several diseases and conditions if the related data is available.

R Packages for Differential Expression Analysis

In this section, we will overview the most popular R packages for this type of analysis:

• limma: Linear Models for Microarray Data (limma) is a powerful and versatile R package that analyzes gene expression data from microarray and, most recently, RNA-seq experiments. It uses linear models to evaluate differential expression and incorporates empirical Bayes methods to improve the stability and reliability of statistical estimates in various scenarios, including experiments with small sample sizes. By far, limma is one of the most popular R packages for this type of analysis.
• DESeq2: This R package is designed for differential gene expression analysis of RNA-seq data. It employs a model-based normalization approach using the negative binomial distribution to account for differences in sequencing depth and other technical biases. DESeq2 provides powerful statistical tests for identifying differentially expressed genes and adjusts for multiple tests to control false discovery rates. It also offers the Variance Stabilizing Transformation (VST) to stabilize variance across mean values, facilitating data visualization and downstream analysis. DESeq2 handles low-count genes effectively and includes various functions for diagnostic and exploratory plots, making it an essential tool for RNA-seq data analysis.
• edgeR: This package analyzes RNA-seq data, focusing on identifying differentially expressed genes. It utilizes the Trimmed Mean of M-values (TMM) normalization method, addressing differences in library sizes across samples. edgeR applies empirical Bayes methods to estimate dispersion and fit statistical models, providing reliable results even with complex experimental designs. The package supports exact tests for simple comparisons and generalized linear models (GLMs) for more intricate studies. Additionally, edgeR includes effective methods for filtering lowly expressed genes, enhancing the accuracy and interpretability of differential expression analysis in RNA-seq data.

Conclusion

R offers a rich ecosystem of packages for microarray and RNA-Seq data analysis, providing comprehensive tools for preprocessing and differential expression analysis. By leveraging these packages, researchers can ensure robust and reproducible results in their transcriptomics studies.