RNA-seq: impact of RNA degradation on transcript quantification.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Assessing whole-genome gene expression using low-quality RNA samples is controversial due to potential biases from non-uniform transcript degradation.
• Degradation may affect different transcripts at varying rates, complicating accurate expression level measurements.
• Low-quality samples are often the only option for research, particularly in fieldwork settings.

Purpose of the Study:

• To quantify the impact of RNA quality variation on gene expression level estimates derived from RNA-sequencing (RNA-seq) data.
• To investigate methods for correcting expression data affected by RNA degradation.

Main Methods:

• Collected RNA-seq expression data from tissue samples with varying degradation levels.
• Included samples spanning the full range of RNA Integrity Number (RIN) values.
• Employed a linear model framework to control for RIN effects.

Main Results:

• Observed significant effects of RNA quality on gene expression measurements across all RIN values.
• Detected a notable loss of library complexity in more degraded RNA samples.
• Standard normalization methods were insufficient to correct for degradation-induced biases.

Conclusions:

• Explicitly modeling RNA Integrity Number (RIN) can correct most degradation effects in gene expression data.
• This approach enables the recovery of biologically relevant signals from degraded RNA samples when RIN is not associated with the variable of interest.
• The findings support the use of degraded RNA samples with appropriate analytical corrections for gene expression studies.