Overview

GiottoDB is a module in the Giotto Suite ecosystem that provides database support for core Giotto functionality through dbverse.

GiottoDB extends Giotto objects with database-backed slots and a persistent database connection.

GiottoDB function support

The table below summarizes GiottoDB-native methods and Giotto functions that are supported for GiottoDB objects.

Category	Function	Description
Object creation	`GiottoDB()`	Create a database-backed GiottoDB object.
Object creation	`as_giottodb()`	Convert a Giotto object to a GiottoDB object.
Object creation	`as_giotto()`	Convert a GiottoDB object back to an in-memory Giotto object.
Saving and loading	`saveGiotto()`	Save a GiottoDB object with its database-backed data.
Saving and loading	`loadGiotto()`	Load a saved GiottoDB object.
Saving and loading	`loadGiottoDB()`	Load a GiottoDB object from a saved GiottoDB project.
Connection management	`dbReconnect()`	Reconnect stale database-backed slots after loading.
Configuration	`GiottoDB-options`	Review global options used by GiottoDB-backed workflows.
Expression objects	`createExprObj()`	Create Giotto expression objects that can contain database-backed matrices.
Expression access	`getExpression()`	Retrieve expression objects or matrices from GiottoDB objects.
Expression processing	`filterGiotto()`	Filter cells and features while preserving database-backed expression data.
Expression processing	`processExpression()`	Run supported expression-processing workflows on dbMatrix-backed expression data.
Expression processing	`processData()`	Run dbMatrix-backed expression processing methods used by Giotto workflows.
Expression processing	`normalizeGiotto()`	Normalize database-backed expression matrices.
Expression processing	`addStatistics()`	Add cell and feature statistics from database-backed matrices.
Expression processing	`calculateHVF()`	Identify highly variable features from database-backed matrices.
Dimension reduction	`runPCA()`	Run PCA using database-backed matrix operations.
Dimension reduction	`runUMAP()`	Run UMAP from supported reduced dimensions.
Clustering	`doLeidenCluster()`	Cluster cells from supported nearest-neighbor graphs or reduced dimensions.
Clustering	`doLeidenClusterIgraph()`	Run Leiden clustering through Giotto’s igraph workflow.
Differential expression	`findMarkers_one_vs_all()`	Run one-versus-all marker testing on supported GiottoDB objects.
Subcellular workflow	`createGiottoPoints()`	Create Giotto point objects backed by dbSpatial where applicable.
Subcellular workflow	`createGiottoPolygon()`	Create Giotto polygon objects backed by dbSpatial where applicable.
Subcellular workflow	`calculateOverlap()`	Compute overlaps between spatial features.
Subcellular workflow	`overlapToMatrix()`	Convert spatial overlaps to matrix-like outputs.
Subcellular workflow	`tessellate()`	Generate square or hexagonal tessellations.
Visualization	`spatPlot2D()`	Create standard Giotto spatial plots for supported GiottoDB objects.
Visualization	`spatInSituPlotPoints()`	Visualize in situ point features from GiottoDB objects.
Visualization	`plotPCA()`	Plot PCA embeddings from GiottoDB objects.
Visualization	`plotUMAP()`	Plot UMAP embeddings from GiottoDB objects.

Performance

Benchmark summaries are available in the benchmarks vignette.