Find bridge anchors between query and extended bridge-reference

Source: R/integration.R

Find a set of anchors between unimodal query and the other unimodal reference using a pre-computed BridgeReferenceSet. This function performs three steps: 1. Harmonize the bridge and query cells in the bridge query reduction space 2. Construct the bridge dictionary representations for query cells 3. Find a set of anchors between query and reference in the bridge graph laplacian eigenspace These anchors can later be used to integrate embeddings or transfer data from the reference to query object using the MapQuery object.

FindBridgeTransferAnchors(
  extended.reference,
  query,
  query.assay = NULL,
  dims = 1:30,
  scale = FALSE,
  reduction = c("lsiproject", "pcaproject"),
  bridge.reduction = c("direct", "cca"),
  verbose = TRUE
)

Arguments

extended.reference

BridgeReferenceSet object generated from PrepareBridgeReference

query

A query Seurat object

query.assay

Assay name for query-bridge integration

dims

Number of dimensions for query-bridge integration

scale

Determine if scale the query data for projection

reduction

Dimensional reduction to perform when finding anchors. Options are:

  • pcaproject: Project the PCA from the bridge onto the query. We recommend using PCA when bridge and query datasets are from scRNA-seq

  • lsiproject: Project the LSI from the bridge onto the query. We recommend using LSI when bridge and query datasets are from scATAC-seq or scCUT&TAG data. This requires that LSI or supervised LSI has been computed for the bridge dataset, and the same features (eg, peaks or genome bins) are present in both the bridge and query.

bridge.reduction

Dimensional reduction to perform when finding anchors. Can be one of:

  • cca: Canonical correlation analysis

  • direct: Use assay data as a dimensional reduction

verbose

Print messages and progress

Value

Returns an AnchorSet object that can be used as input to TransferData, IntegrateEmbeddings and MapQuery.