BCRgt

BCRgt performs genotype calling in tumor samples exhibiting copy number alterations (CNAs) to improve accuracy of variant detection for Genome-Wide Association Studies (GWAS) and cancer genomics.

Key Features:

• Bayesian enhancement: Incorporates a Bayesian layer into a cluster regression model to probabilistically model genotype calls in regions affected by CNAs.
• Training step integration: Includes a training step that increases concordance with HapMap genotype calls even with limited sample sizes.
• Improved accuracy in CNA regions: Enhances genotyping accuracy in regions of DNA copy loss and provides modest improvements in copy number gain regions relative to the Bayesian Robust Linear Model with Mahalanobis distance classifier (BRLMM).

Scientific Applications:

• Cancer genomics: Enables accurate genotype calls in tumor samples with CNAs to support analyses of genetic variants related to cancer development, progression, and treatment.
• Genome-Wide Association Studies (GWAS): Facilitates reliable association testing by providing accurate genotypes for tumor-derived samples affected by CNAs.

Methodology:

Implements a Bayesian cluster regression framework with an explicit Bayesian layer and a training step to probabilistically model genotype calls in regions affected by CNAs.