Inference of alternative splicing from RNA-Seq data with probabilistic splice graphs.

Area of Science:

• Molecular Biology
• Bioinformatics
• Genomics

Background:

• Alternative splicing generates diverse transcripts from single genes, crucial for eukaryotic cell function.
• RNA sequencing (RNA-Seq) is a powerful tool for transcript analysis, adaptable to unknown gene structures.
• Analyzing RNA-Seq data for genes with numerous alternative transcripts presents challenges in efficiency, identifiability, and representation.

Purpose of the Study:

• To develop novel RNA-Seq models and inference algorithms for analyzing complex transcriptomes.
• To address efficiency, identifiability, and representation issues in RNA-Seq data analysis.
• To enable accurate quantification and differential processing detection in the presence of complex alternative splicing.

Main Methods:

• Probabilistic splice graphs were utilized to model alternative splicing events.
• Development of associated inference algorithms for RNA-Seq data analysis.
• Mathematical proofs demonstrating model identifiability.

Main Results:

• The proposed models based on probabilistic splice graphs effectively alleviate existing analysis challenges.
• The inference methods demonstrate efficiency and accuracy in transcript quantification.
• Differential processing detection using these methods is also shown to be efficient and accurate.

Conclusions:

• The developed probabilistic splice graph models offer a robust solution for analyzing complex RNA-Seq data.
• The inference algorithms provide accurate and efficient quantification and differential processing detection.
• This work advances the analysis of alternative splicing and transcript diversity in eukaryotic gene expression.