Errors in RNA-Seq quantification affect genes of relevance to human disease

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• RNA sequencing (RNA-Seq) is a standard for gene expression measurement, crucial for human disease studies.
• Accurate gene expression quantification depends on unique read information for mapping to genomic references.
• Current methods face challenges with ambiguous sequence reads, potentially leading to inaccurate gene expression estimates.

Purpose of the Study:

• To identify genes with underestimated expression using common RNA-Seq quantification methods.
• To propose and validate a novel two-stage analysis for handling ambiguous RNA-Seq reads.
• To demonstrate the utility of this method in extracting biological insights from previously discarded data.

Main Methods:

• Application of 12 common gene expression quantification methods to RNA-Seq data.
• Development of a two-stage analysis to group multi-mapped or ambiguous reads.
• Testing the proposed method on a mouse cancer study dataset.

Main Results:

• Hundreds of genes exhibit underestimated expression across multiple standard RNA-Seq quantification methods.
• Many of these affected genes are linked to human diseases and belong to gene families.
• The two-stage analysis successfully assigned ambiguous reads to gene groups, extracting biological signals from discarded data.

Conclusions:

• RNA-Seq exhibits limitations in accurately measuring expression for a significant number of human genes.
• These limitations disproportionately affect gene families and genes implicated in human diseases.
• A novel analytical approach can leverage previously discarded RNA-Seq data to provide biologically relevant group-level expression insights.