lme4qtl

lme4qtl extends lme4 to enable quantitative trait locus (QTL) mapping by fitting linear mixed models (LMMs) for related individuals using custom covariance structures.

Key Features:

• Custom Covariance Matrices: Allows random effects to be defined using user-specified covariance matrices, including multiple covariance matrices to model complex levels of relatedness.
• Parameter Constraints: Implements parameter constraints to enforce biologically meaningful limits during model estimation.
• Integration with lme4: Integrates with the lme4 framework and its companion packages to leverage established LMM functionality.
• Efficiency in Sparse Covariance Scenarios: Optimizes computations for sparse covariance matrices to improve performance in large family-structured genetic studies.

Scientific Applications:

• Linkage Analysis: Supports linkage studies by modeling relatedness among individuals within LMMs.
• Association Studies and GWAS: Enables association analyses and genome-wide association study (GWAS) scenarios that require modeling of various relatedness levels.
• Family-based QTL Mapping (GAIT2): Applied to family-based datasets such as the Genetic Analysis of Idiopathic Thrombophilia 2 (GAIT2) project to demonstrate practical utility.

Methodology:

Fits linear mixed models via lme4, supports custom and multiple covariance matrices for random effects, enforces parameter constraints, and applies computational optimizations for sparse covariance matrices.