Function compute_highly_variable_genes

src/memory/processing/hvg/mod.rs:156–168

Compute highly variable genes using the specified method and parameters. This is the main entry point for HVG detection. It analyzes the mean-variance relationship in the expression data to identify genes with biological variability above technical noise. ## Algorithm Overview 1. **Calculate Statistics**: Compute mean and variance for each gene 2. **Model Relationship**: Fit mean-variance relat

(
    adata: &IMAnnData,
    params: Option<HVGParams>,
)

/// compute_highly_variable_genes(&adata, Some(params))?;
/// ```
pub fn compute_highly_variable_genes(
    adata: &IMAnnData,
    params: Option<HVGParams>,
) -> anyhow::Result<()> {
    let params = params.unwrap_or_default();
    let x = adata.x();

    match params.flavor {
        FlavorType::Seurat => compute_seurat_hvg(adata, &x, params),
        FlavorType::CellRanger => compute_cell_ranger_hvg(adata, &x, params),
        FlavorType::SVR => compute_svr_hvg(adata, &x, params),
    }
}

/// Post-process dispersion statistics for Seurat method to handle edge cases.
///