strum

strum implements structural equation modeling for genetic analysis of general pedigree data in R, accounting for measurement error, familial correlations, and ascertainment bias.

Key Features:

• General pedigree support: Implements SEM tailored to general pedigree data including familial correlation structures.
• Confirmatory factor analysis: Integrates confirmatory factor analysis within SEM frameworks for latent variable measurement modeling.
• Measurement error modeling: Explicitly models measurement error within structural and measurement components.
• Ascertainment bias handling: Includes mechanisms to account for ascertainment issues common in genetic studies.
• Genetic association analysis: Supports assessment of relationships between genetic variants and traits within SEM.
• Linkage analysis: Provides modeling capabilities for identifying genomic regions associated with phenotypes.
• Polygenic effects: Models the influence of multiple genes (polygenic effects) on traits.
• Shared environment: Models shared environmental factors contributing to familial correlations.
• Pedigree simulation: Includes tools to simulate pedigree data for hypothesis testing and validation.
• Model visualization: Offers visualization of SEM models to aid interpretation of complex genetic structures.
• Simulation-based evaluation: Has been evaluated via simulation studies reporting parameter estimate accuracy and coverage.

Scientific Applications:

• Genetic association testing: Assess relationships between genetic variants and phenotypic traits within pedigrees.
• Linkage mapping: Identify genomic regions linked to phenotypes using pedigree-based SEM.
• Polygenic architecture analysis: Quantify contributions of multiple genes to complex trait variation.
• Familial environment modeling: Partition genetic and shared environmental contributions to trait covariance.
• Joint measurement and structural modeling: Combine measurement models (latent variables) with structural models in pedigree analyses.
• Method validation: Use simulated pedigree data to evaluate estimator properties and model performance.

Methodology:

Implements confirmatory factor analysis and general structural equation modeling adapted to general pedigree data, modeling familial correlations and ascertainment bias, and includes pedigree simulation with simulation-based evaluation of parameter estimates and coverage.