Data quality aware analysis of differential expression in RNA-seq with NOISeq R/Bioc package

TLDR

RNA‑seq analysis increasingly emphasizes experimental design, bias removal, accurate quantification, and false‑positive control, and NOISeq provides a comprehensive resource to meet these data‑aware needs. The study introduces the NOISeq R‑package for quality control and analysis of RNA‑seq count data. The package offers diagnostic tools to monitor quality, guide pre‑processing decisions, and enhance downstream analysis. NOISeqBIO efficiently controls false discoveries in replicated experiments and outperforms state‑of‑the‑art methods.

Abstract

As the use of RNA-seq has popularized, there is an increasing consciousness of the importance of experimental design, bias removal, accurate quantification and control of false positives for proper data analysis. We introduce the NOISeq R-package for quality control and analysis of count data. We show how the available diagnostic tools can be used to monitor quality issues, make pre-processing decisions and improve analysis. We demonstrate that the non-parametric NOISeqBIO efficiently controls false discoveries in experiments with biological replication and outperforms state-of-the-art methods. NOISeq is a comprehensive resource that meets current needs for robust data-aware analysis of RNA-seq differential expression.