MTG2

MTG2 implements accelerated REML-based estimation of genetic and environmental variance and covariance from genome-wide SNPs using multivariate linear mixed models for analysis of complex traits.

Key Features:

• Accelerated computation: Provides faster computation compared to standard REML-based methods for variance component estimation.
• Scalability: Optimized for large-scale genomic datasets using genome-wide SNPs.
• Multivariate analysis: Supports multivariate linear mixed model analyses to estimate covariances among traits.
• Multi-trait models: Implements multi-trait modeling to investigate genetic covariance and correlation across traits.
• Random regression models: Supports random regression models for studying reaction norms and genotype-by-environment variation.
• Cross-species applicability: Demonstrated applicability to both mice and human datasets.

Scientific Applications:

• Genetic analysis of complex traits: Estimates genetic and environmental variance and covariance from genome-wide SNPs within a linear mixed model framework.
• Multi-Trait Models: Investigates how different traits co-vary genetically using multivariate models.
• Random Regression Models: Studies reaction norms to understand how genetic expression changes across environments or conditions.

Methodology:

MTG2 employs a novel algorithm to optimize REML-based analyses of multivariate linear mixed models using genome-wide SNPs to obtain estimates of variance components.