Skip to contents

Run Scissor phenotype-associated cell selection

Source: R/RunScissor.R
RunScissor.Rd

Run Scissor phenotype-associated cell selection

Usage

RunScissor(
  srt,
  bulk_dataset,
  phenotype = NULL,
  condition.by = NULL,
  positive = NULL,
  assay = NULL,
  bulk_assay = "counts",
  layer = "counts",
  features = NULL,
  family = c("gaussian", "binomial", "cox"),
  backend = c("cpp", "r"),
  alpha = NULL,
  cutoff = 0.2,
  tag = NULL,
  graph = NULL,
  dims = 1:10,
  nfeatures = 2000,
  seed = 123,
  prefix = "Scissor",
  tool_name = "Scissor",
  store_inputs = FALSE,
  verbose = TRUE
)

Arguments

srt

A Seurat object containing single-cell expression data.

bulk_dataset

A bulk expression matrix-like object or a SummarizedExperiment. Rows are genes and columns are bulk samples.

phenotype

Phenotype annotation for bulk samples. For family = "binomial", character or factor values are converted to 0/1 according to positive. If NULL and bulk_dataset is a SummarizedExperiment, condition.by is used.

condition.by

Column in colData(bulk_dataset) used as phenotype when phenotype = NULL.

positive

Positive phenotype level for binomial Scissor. If NULL, the second sorted level is used.

assay

Which assay to use. If NULL, the default assay of the Seurat object will be used. When the object also contains ChromatinAssay, the default assay and additional ChromatinAssay will be preprocessed sequentially.

bulk_assay

Assay name used when bulk_dataset is a SummarizedExperiment.

layer

Assay layer used from srt.

features

Optional genes used before intersecting bulk and single-cell features.

family

Regression family passed to Scissor.

backend

Scissor backend. "r" calls the upstream package path and "cpp" uses the optimized scop path. Legacy aliases "original" and "scop" are accepted as "r" and "cpp", respectively.

alpha

Scissor alpha search values. If NULL, Scissor's default grid is used.

cutoff

Maximum selected-cell fraction used by Scissor's alpha search.

tag

Optional phenotype labels for printed/stored summaries.

graph

Existing Seurat graph used by the "cpp" backend. If NULL, an assay SNN graph or standard_scop() graph is reused when present, otherwise a temporary Scissor-like graph is built.

dims

PCA dimensions used when a temporary graph is built.

nfeatures

Number of variable features used when a temporary graph is built.

seed

Random seed used by Scissor's alpha loop.

prefix

Prefix for metadata column names.

tool_name

Name of the srt@tools entry.

store_inputs

Whether to store Scissor regression inputs in srt@tools. Default is FALSE to avoid large objects.

verbose

Whether to print the message. Default is TRUE.

Value

A Seurat object with Scissor status and coefficient columns in metadata and a Scissor result bundle in srt@tools[[tool_name]].

References

Sun, D. et al. Identifying phenotype-associated subpopulations by integrating bulk and single-cell sequencing data. Nature Biotechnology (2021). doi:10.1038/s41587-021-01091-3

See also

ScissorPlot

Examples

data(panc8_sub)
data(islet_bulk)
panc8_sub <- standard_scop(panc8_sub, verbose = FALSE)
panc8_sub <- RunScissor(
  panc8_sub,
  bulk_dataset = islet_bulk,
  condition.by = "condition",
  positive = "bfa",
  family = "binomial",
  features = head(
    intersect(
      rownames(panc8_sub),
      rownames(SummarizedExperiment::assay(islet_bulk, "counts"))
    ), 1000
  ),
  alpha = 0.2,
  cutoff = 0.5
)

ScissorPlot(
  panc8_sub,
  xlab = "UMAP_1",
  ylab = "UMAP_2"
)