MIXnorm: normalizing RNA-seq data from formalin-fixed paraffin-embedded samples.

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• RNA-sequencing (RNA-seq) analysis of formalin-fixed paraffin-embedded (FFPE) tissues is valuable for transcriptome-wide studies.
• Normalization of FFPE RNA-seq data is crucial but underexplored, as existing methods are optimized for fresh-frozen samples.
• Suboptimal normalization can introduce biases, affecting the accuracy of gene expression measurements.

Purpose of the Study:

• To develop and evaluate a novel normalization method specifically designed for FFPE RNA-seq data.
• To address the limitations of existing normalization techniques when applied to FFPE samples.
• To improve the reliability and accuracy of gene expression analysis from FFPE tissues.

Main Methods:

• Proposed MIXnorm, a normalization method utilizing a two-component mixture model.
• Modeled non-expressed genes with zero-inflated Poisson distributions and expressed genes with truncated normal distributions.
• Developed a nested EM algorithm for efficient computation of maximum likelihood estimates without numerical optimization.

Main Results:

• MIXnorm demonstrated significant improvements compared to commonly used RNA-seq normalization methods.
• The nested EM algorithm provided a computationally efficient and easy-to-implement solution.
• Evaluations through simulations and cancer studies confirmed the efficacy of MIXnorm for FFPE RNA-seq data.

Conclusions:

• MIXnorm offers a robust and efficient normalization strategy for FFPE RNA-seq data.
• The method enhances the quality of transcriptome analysis from FFPE samples.
• MIXnorm represents a significant advancement in processing and analyzing FFPE-derived gene expression data.