API reference¶
The API is organized by the role each function plays in a SampleDisco run. Functions shared between modalities (cell typing, sample embedding, and the whole downstream module) are documented once and cross-linked.
The config-driven wrapper is the supported entry point
Most users should not call these functions directly. The primary interface is the YAML-config wrapper — sampledisco -m complex --config <config.yaml> (equivalently python -m sampledisco.cli -m complex --config <config.yaml>), which dispatches sampledisco.wrapper.wrapper.wrapper(**config). It orchestrates preprocess → cell embedding → cell typing → sample embedding → downstream for each enabled modality. The functions below are the building blocks it calls.
Public import paths
The import root is sampledisco (NOT genodistance). Subpackage __init__ files are empty, so import each public function from its concrete module — e.g. from sampledisco.preparation.rna_preprocess_cpu import preprocess, from sampledisco.sample_embedding import compute_sample_embedding. The two top-level re-exports are sampledisco.wrapper and sampledisco.compute_sample_embedding.
Categories¶
| Section | When you reach for it | Key functions |
|---|---|---|
| RNA | scRNA-seq preprocessing and clustering. | preprocess, cell_types |
| ATAC | scATAC-seq preprocessing and cell typing. | preprocess (ATAC build), cell_types_atac |
| Multi-omics | GLUE integration, joint cell typing, multi-omics sample embedding, embedding merge. | multiomics_preparation, cell_types_multiomics, compute_sample_embedding |
| Shared | The core sample embedding — reused by all modalities. | compute_sample_embedding |
| Autotune | Parameter selection — optimizes the RMD-weight alpha (composition vs displacement). | run_autotune |
| Downstream | Everything that runs off the sample embedding: distance, trajectory, DGE, clustering, visualization. | sample_distance, CCA_Call, cca_pvalue_test, TSCAN, run_trajectory_gam_differential_gene_analysis, cluster, proportion_test, raisinfit, run_pairwise_tests, run_dimension_association_analysis, visualization, visualize_multimodal_embedding |
If you are unsure which function to call next, the tutorials walk through the end-to-end order: RNA, ATAC, Multi-omics, Downstream.