Tutorials

Learn gsuite through worked examples in R.

Choose a complete analysis to learn how gsuite is used. Detailed options and data-preparation steps are collected under Guides; results used to assess the software are under Software validation.

For the standalone package tutorials, install the required packages from the source-package repository. Pedigree and root-integration examples require an existing development installation; greml/gsolve R packaging is deferred.

Standalone package tutorials

One simulated dataset with 20,000 individuals, 50,000 markers and three traits is followed across the seven public R packages. Start with the study design and reproducibility guide, then follow each package’s worked analysis. The pages explain inputs, result columns, diagnostics, alternative methods and which outputs were actually generated. Displayed results come from one representative run with a four-thread limit; they are not a performance comparison.

Package	Tutorial	Main result
gbase	Prepare genotypes and LD	Glist, reusable sparse LD and aligned statistics
glma	Association analysis	Marker tests and LD-aware VEGAS gene tests
gbayes	Bayesian genomic regression	Posterior marker effects and scoring weights
gcorr	Heritability and genetic correlation	LDSC estimates and uncertainty
gmap	Regional fine-mapping	Inclusion probabilities, credible sets and diagnostics
gscore	Polygenic scoring	Individual total and marker-set scores
gsea	Gene and pathway analysis	Pathway tests and PoPS-style gene ranking

Pedigree analysis

Development example: requires the root R integration, which is not distributed in the public source repository.

Fit pedigree models with R data frames

Start with a single-trait animal model, then use the same workflow for multitrait and longitudinal analyses. The tutorial shows model specification, fitting, summaries and predictions.

Genomic analysis

Follow one dataset across six genomic modules

Begin with genotypes and phenotypes, run GWAS, and reuse the resulting summary statistics and LD for Bayesian regression, genetic correlation, mapping, scoring and pathways. This separate qualification example uses a 10,000-individual, 50,000-marker simulation.

Association analysis

Run association analyses with glma

Reuse the included artificial genotype panel for quantitative, conditional, binary and LOCO scans, then aggregate the marker results with VEGAS. The tutorial explains the returned tables, model information and diagnostics.

Summary-statistic analysis

Work from a saved LD reference to genetic analysis

Prepare and reload LD, analyse three traits, then inspect genome-wide, annotation, regional and conditional results. The gcorr guide explains the method choices and assumptions.

Working with an assistant

The pedigree-based walkthroughs use the development R integration.

From a question to a reproducible analysis

Work through data, model specification, execution, diagnostics and a report.

From a question to a supported model

See how a new analysis request can become a focused, checked extension to the existing computational components.