Markdown tutorials with executed code and embedded output figures — no website build needed, readable on GitHub, in an editor, or by a coding agent. Every output shown was produced by the code above it, on the named public dataset, with the LARIS version stated in each file.
Written for LARIS v0.12.0. If you are reporting p-values, read tutorial 07 — significance testing changed in this release.
| Tutorial | Dataset | Covers |
|---|---|---|
| 01 — Core pipeline | Slide-tags human tonsil (Zenodo 10.5281/zenodo.19981287) | prepareLRInteraction, runLARIS, spatial / dot / network plots, defaults & reference values |
| 02 — Compare conditions (aggregate) | Kuppe MI Visium (cellxgene) | compareLARIS from per-sample results tables, subjects vs slices, Fisher detection route, volcano |
| 03 — Compare at matched cell states | MERFISH gut atlas (Dryad 10.5061/dryad.p5hqbzm0z) | matched estimator, buildJointEmbedding (harmony / PIASO GDR / PCA), per-subject profiles, combineComparisons |
| 04 — Cytome guide | tonsil | the on-disk workflow end to end: LR cytome, streaming COSG, bit-identical guarantee |
| 05 — Tissue image overlay | Xenium Prime 5K human skin (10x) | H&E overlay at single-cell resolution, registering a post-hoc image, 5K-panel LR analysis |
| 06 — Two-variable analysis | MERFISH gut atlas | crossed cell-type x region labels (PIASO getCrossCategories), region-resolved comparison |
| 07 — Statistical significance | Slide-tags human tonsil | prepareLRBackground, what the p-value tests, choosing n_matched_genes, reusing one background across analyses |
For the steps upstream of LARIS on spatial data, QC, clustering, annotation, RNA regulon analysis and visualization, see PIASO's spatial tutorials.
LARIS is part of the PIASO ecosystem — the tutorials use PIASO's figure styling and mention its INFOG/GDR options where they fit — but PIASO is not a required dependency: every LARIS step runs without it, and optional packages raise an informative install pointer only when a feature that needs them is requested.