A random coefficients model for regional co-expression associated with DNA copy number

Area of Science:

• Genomics
• Statistical Genetics
• Bioinformatics

Background:

• Regional co-expression, where nearby genes show similar expression patterns, is a key genomic phenomenon.
• DNA copy number aberrations (CNAs) are implicated as a potential driver of regional co-expression.
• Understanding the relationship between CNAs and gene expression is crucial for deciphering genomic regulation.

Purpose of the Study:

• To develop a statistical model that explains regional co-expression using DNA copy number information.
• To differentiate the effects of CNAs from other factors contributing to co-expression.
• To assess the homogeneity of CNA effects across different genes.

Main Methods:

• A random coefficients model is proposed to link DNA copy number data with regional co-expression.
• The model incorporates a correlated error structure to account for residual co-expression from other sources.
• Computationally efficient and consistent estimation methods are presented for high-dimensional genomic data.
• Robustification techniques are suggested for the estimation procedure.

Main Results:

• The proposed model allows for the estimation of parameters reflecting the influence of CNAs on gene expression.
• It enables the assessment of whether CNA effects are shared across genes or gene-specific.
• The methodology is demonstrated through two application examples investigating CNA-regional co-expression associations.

Conclusions:

• The developed random coefficients model provides a robust framework for analyzing regional co-expression in the context of DNA copy number variations.
• The approach facilitates the investigation of shared and heterogeneous effects of CNAs on gene expression patterns.
• This methodology offers valuable insights into the genomic mechanisms underlying regional gene expression regulation.