A BOOTSTRAP MODEL COMPARISON TEST FOR IDENTIFYING GENES WITH CONTEXT-SPECIFIC PATTERNS OF GENETIC REGULATION.

Area of Science:

• Genetics and Genomics
• Systems Biology
• Bioinformatics

Background:

• Genetic variation significantly influences gene expression, impacting complex traits.
• Existing research confirms differential gene expression across human tissues and cell types.
• A critical gap exists in tools for identifying genes with differentially regulated expression.

Approach:

• Introduced Differential Regulation Analysis by Bootstrapping (DRAB), a novel gene-based method.
• DRAB employs elastic net to build context-specific genetic regulation models.
• A bootstrap-based model comparison test assesses the equivalency of these models, accounting for training variability.

Key Points:

• DRAB effectively determines if population-level genetic regulation models exhibit equal predictive performance.
• The method accounts for variability in feature selection and model training, improving accuracy.
• Validated on Genotype-Tissue Expression (GTEx) Project data, DRAB demonstrated biological relevance.

Conclusions:

• DRAB possesses sufficient power to detect tissue-specific regulatory profiles while controlling false positives.
• This framework aids in prioritizing differentially regulated genes.
• Enables enhanced understanding of the genetic architecture underlying molecular phenotypes.