Analyzing allele specific RNA expression using mixture models.

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Allele-specific RNA expression (AEI) analysis offers insights into gene regulation.
• RNA-sequencing (RNA-Seq) enables transcriptome-wide AEI measurement at heterozygous single nucleotide polymorphisms (SNPs).
• Current methods often assume a biologically unrealistic negative correlation between reference and variant allele reads.

Purpose of the Study:

• To develop a novel strategy for AEI analysis using RNA-Seq data.
• To address limitations of existing methods in modeling allelic expression imbalance.
• To identify AEI without assuming a specific correlation direction between allele reads.

Main Methods:

• Proposed a new strategy for AEI analysis in RNA-Seq data.
• Grouped comparable SNPs based on similar total allelic read counts.
• Applied a mixture of folded Skellam distributions to identify AEI signal in comparable SNP groups.

Main Results:

• Identified numerous instances of moderate to strong AEI at heterozygous SNPs in human brain tissues.
• Demonstrated the method's applicability to RNA-Seq data.
• Highlighted findings for SLC1A3 mRNA as a case study.

Conclusions:

• The folded Skellam mixture model effectively detects AEI by comparing reference and variant allele reads.
• This model is suitable for genes with few heterozygous SNPs and handles over-dispersed read counts.
• The approach provides a more flexible and biologically plausible framework for AEI analysis.