Skip to contents

Compute LISI scores on a Seurat object

Source: R/LISI.R
RunLISI.Rd

Compute per-cell Local Inverse Simpson's Index (LISI) scores from a dimensional reduction and store them in the meta.data and tools slots of a Seurat object.

Usage

RunLISI(
  srt,
  reductions = NULL,
  reduction = NULL,
  dims = NULL,
  label_colnames = NULL,
  prefix = NULL,
  tool_name = NULL,
  perplexity = 30,
  tol = 1e-05,
  max_iter = 50,
  knn_algorithm = c("auto", "brute_force", "clustered"),
  cores = NULL,
  max_dense_bytes = Inf,
  overwrite = TRUE,
  verbose = TRUE
)

Arguments

srt

A Seurat object.

reductions

Character vector of dimensional reductions used to compute LISI. If NULL, DefaultReduction() is used.

reduction

Deprecated alias of reductions.

dims

Dimensions to use from the reduction. Default is NULL, which uses all available dimensions.

label_colnames

Character vector of metadata columns used for LISI. If NULL, RunLISI() will try to use srt@misc[["integration_batch"]].

prefix

Prefix used for the stored LISI metadata columns. If NULL, the reduction names are used.

tool_name

Name used to store detailed results in srt@tools. Default is "LISI" when multiple reductions are provided, otherwise paste0(prefix, "_LISI").

perplexity

Effective neighborhood size.

tol

Tolerance used in the binary search for the target perplexity.

max_iter

Maximum number of binary-search iterations.

knn_algorithm

Exact nearest-neighbor strategy passed to thisutils::compute_lisi().

cores

Number of LISI C++ worker threads. NULL (the default) lets thisutils::compute_lisi() select the available hardware threads.

max_dense_bytes

Maximum estimated bytes allowed for LISI's dense input and C++ copy. Default is Inf, which preserves unrestricted behavior.

overwrite

Whether to overwrite existing metadata columns.

verbose

Whether to print the message. Default is TRUE.

Value

A modified Seurat object.

See also

thisutils::compute_lisi, IntegrationBenchmarkPlot, RunIntegrationBenchmark

Examples

data(panc8_sub)
set.seed(1)
demo_embedding <- matrix(
  stats::rnorm(ncol(panc8_sub) * 5),
  nrow = ncol(panc8_sub),
  dimnames = list(colnames(panc8_sub), paste0("DEMO_", 1:5))
)
panc8_sub[["demo"]] <- SeuratObject::CreateDimReducObject(
  embeddings = demo_embedding,
  key = "DEMO_",
  assay = SeuratObject::DefaultAssay(panc8_sub)
)
names(panc8_sub@reductions)
#> [1] "demo"

panc8_sub <- RunLISI(
  panc8_sub,
  reductions = "demo",
  label_colnames = "tech",
  perplexity = 10
)
#> ℹ [2026-08-30 05:16:13] Compute LISI scores from reduction "demo"
#> ✔ [2026-08-30 05:16:13] Stored LISI scores in metadata: "demo_tech_LISI"
IntegrationBenchmarkPlot(panc8_sub, plot_type = "box")