2016/09/04 by Ciaran Evans, Evans, Ciaran, Johanna Hardin +3 · 1 citation
Biochemistry, Genetics and Molecular Biology · #FOS: Biological sciences #Gene expression and cancer classification #Genomics (q-bio.GN) #Molecular Biology Techniques and Applications #RNA Research and Splicing
paper · pdf · doi:10.48550/arxiv.1609.00959
openalex publication_date 2016/09/04 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
RNA-Seq is a widely-used method for studying the behavior of genes under\ndifferent biological conditions. An essential step in an RNA-Seq study is\nnormalization, in which raw data are adjusted to account for factors that\nprevent direct comparison of expression measures. Errors in normalization can\nhave a significant impact on downstream analysis, such as inflated false\npositives in differential expression analysis. An under-emphasized feature of\nnormalization is the assumptions upon which the methods rely and how the\nvalidity of these assumptions can have a substantial impact on the performance\nof the methods. In this paper, we explain how assumptions provide the link\nbetween raw RNA-Seq read counts and meaningful measures of gene expression. We\nexamine normalization methods from the perspective of their assumptions, as an\nunderstanding of methodological assumptions is necessary for choosing methods\nappropriate for the data at hand. Furthermore, we discuss why normalization\nmethods perform poorly when their assumptions are violated and how this causes\nproblems in subsequent analysis. To analyze a biological experiment,\nresearchers must select a normalization method with assumptions that are met\nand that produces a meaningful measure of expression for the given experiment.\n