bio-causal-genomics-genomic-sem - Model latent genetic architecture with GenomicSEM
Fits GenomicSEM models to GWAS summary statistics, runs multivariate genetic analyses, evaluates model fit, and compares results with MTAG.
Tags
Updated: 2026-10-16Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Compute LDSC covariance matrices
- Fit common-factor models
- Fit user-specified SEMs
- Run common-factor GWAS
- Test Q_SNP heterogeneity
- Run multivariate Wald tests
- Partition factor heritability
- Compare results with MTAG
- Identify Heywood cases
- Verify model fit
Inputs
- Munged GWAS summary statistics
- Lavaan model syntax
- Functional annotations
- MTAG analysis results
Outputs
- Genetic covariance matrix
- Sampling covariance matrix
- Model parameter estimates
- Model fit indices
- SNP-level association statistics
- Q_SNP heterogeneity statistics
- Partitioned heritability estimates
- MTAG comparison results
Requirements
- R 4.4 or later
- GenomicSEM 0.0.5 from GitHub
- lavaan 0.6.19
- LDSC Python 3 installation
- LDSC commit 1f09cf0
- baselineLD_v2.2 annotations
- MTAG 1.0.8 or later for MTAG analyses