Practical RNA-Seq with Spike-Ins: A Bench-to-Bioinformatics Guide

Abstract

RNA-Sequencing (RNA-Seq) is a powerful, high-throughput technology for genome-wide transcriptional analysis, widely used to investigate gene expression changes in plants. A critical step in RNA-Seq data analysis is normalization, which enables accurate comparisons of gene expression between samples by reducing the effects of technical and biological factors that can confound expression measurements. Traditional normalization approaches, such as those implemented in DESeq2's median-of-ratios and edgeR's Trimmed Mean of M-values, operate under the assumption that the total transcript levels or the expression of most genes are stable across samples. However, this assumption is often violated, particularly in plant studies, where treatments, genotypes, or even time-of-sampling differences can affect overall transcription efficiency or RNA stability, producing unequal transcript abundance (Coate and Doyle, Genome Biol Evol 2:534-546, 2010; Laosuntisuk et al. Plant J 118:1241-1257, 2024; Forsythe et al. Proc Natl Acad Sci USA 119:e2204187119, 2022; Wang et al. New Phytol 230:1985-2000, 2021). Such violations can result in inappropriate scaling of expression levels, leading to incorrect identification of differentially expressed genes (DEGs). Exogenous RNA spike-ins offer a robust solution for controlling variations in total mRNA levels and technical biases. This chapter presents practical protocols for integrating exogenous RNA spike-ins at the bench and in downstream computational analysis. We show how to calculate spike-in amounts, implement spike-in-based normalization strategies, and contrast these with distribution-based methods. The goal of this chapter is to lower the barriers to adopting RNA spike-ins for plant transcriptional analysis, thereby improving the accuracy, specificity, and sensitivity of DEG calling in RNA-Seq experiments to ensure more reliable and interpretable biological conclusions.