AXOLOTL: an accurate method for detecting aberrant gene expression in rare diseases using coexpression constraints

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Aberrant transcription events are key to understanding rare diseases and prioritizing causative genes for diagnosis.
• Accurate identification of disease-causing genes remains a significant challenge in clinical genetics.

Purpose of the Study:

• To introduce AXOLOTL, a novel ensemble method for detecting aberrant gene expression in RNA sequencing data.
• To improve the identification and prioritization of pathogenic variants in rare disease patients.

Main Methods:

• AXOLOTL is an ensemble method that incorporates coexpression constraints to account for gene correlations.
• The method was evaluated on diverse RNA-seq datasets, including healthy cohorts and specific rare disease cohorts (mitochondrial diseases, collagen VI-related dystrophy).

Main Results:

• AXOLOTL demonstrated superior performance in identifying aberrant gene expression events.
• The method successfully identified aberrant gene expression in neurological disorder cases, aiding in pathogenic variant prioritization.

Conclusions:

• AXOLOTL offers a robust approach for analyzing aberrant gene expression in rare diseases.
• The tool facilitates enhanced diagnostic accuracy by improving gene prioritization.