clustify: Compare scRNA-seq data to reference data.

In NCBI-Hackathons/RClusterCT: Classifier for Single-cell RNA-seq Using Cell Clusters

Description

Compare scRNA-seq data to reference data.

Usage

clustify(input, ...)

## Default S3 method:
clustify(
  input,
  ref_mat,
  metadata = NULL,
  cluster_col = NULL,
  query_genes = NULL,
  n_genes = 1000,
  per_cell = FALSE,
  n_perm = 0,
  compute_method = "spearman",
  verbose = TRUE,
  lookuptable = NULL,
  rm0 = FALSE,
  obj_out = TRUE,
  seurat_out = TRUE,
  rename_prefix = NULL,
  threshold = "auto",
  low_threshold_cell = 0,
  exclude_genes = c(),
  if_log = TRUE,
  organism = "hsapiens",
  plot_name = NULL,
  ...
)

## S3 method for class 'Seurat'
clustify(
  input,
  ref_mat,
  cluster_col = NULL,
  query_genes = NULL,
  n_genes = 1000,
  per_cell = FALSE,
  n_perm = 0,
  compute_method = "spearman",
  use_var_genes = TRUE,
  dr = "umap",
  seurat_out = TRUE,
  obj_out = TRUE,
  threshold = "auto",
  verbose = TRUE,
  rm0 = FALSE,
  rename_prefix = NULL,
  exclude_genes = c(),
  metadata = NULL,
  organism = "hsapiens",
  plot_name = NULL,
  ...
)

## S3 method for class 'SingleCellExperiment'
clustify(
  input,
  ref_mat,
  cluster_col = NULL,
  query_genes = NULL,
  per_cell = FALSE,
  n_perm = 0,
  compute_method = "spearman",
  use_var_genes = TRUE,
  dr = "umap",
  seurat_out = TRUE,
  obj_out = TRUE,
  threshold = "auto",
  verbose = TRUE,
  rm0 = FALSE,
  rename_prefix = NULL,
  exclude_genes = c(),
  metadata = NULL,
  organism = "hsapiens",
  plot_name = NULL,
  ...
)

Arguments

input	single-cell expression matrix or Seurat object
...	additional arguments to pass to compute_method function
ref_mat	reference expression matrix
metadata	cell cluster assignments, supplied as a vector or data.frame. If data.frame is supplied then cluster_col needs to be set. Not required if running correlation per cell.
cluster_col	column in metadata that contains cluster ids per cell. Will default to first column of metadata if not supplied. Not required if running correlation per cell.
query_genes	A vector of genes of interest to compare. If NULL, then common genes between the expr_mat and ref_mat will be used for comparision.
n_genes	number of genes limit for Seurat variable genes, by default 1000, set to 0 to use all variable genes (generally not recommended)
per_cell	if true run per cell, otherwise per cluster.
n_perm	number of permutations, set to 0 by default
compute_method	method(s) for computing similarity scores
verbose	whether to report certain variables chosen and steps
lookuptable	if not supplied, will look in built-in table for object parsing
rm0	consider 0 as missing data, recommended for per_cell
obj_out	whether to output object instead of cor matrix
seurat_out	output cor matrix or called seurat object (deprecated, use obj_out instead)
rename_prefix	prefix to add to type and r column names
threshold	identity calling minimum correlation score threshold, only used when obj_out = TRUE
low_threshold_cell	option to remove clusters with too few cells
exclude_genes	a vector of gene names to throw out of query
if_log	input data is natural log, averaging will be done on unlogged data
organism	for GO term analysis, organism name: human - 'hsapiens', mouse - 'mmusculus'
plot_name	name for saved pdf, if NULL then no file is written (default)
use_var_genes	if providing a seurat object, use the variable genes (stored in [e-mail]) as the query_genes.
dr	stored dimension reduction

Value

single cell object with identity assigned in metadata, or matrix of correlation values, clusters from input as row names, cell types from ref_mat as column names

Examples

# Annotate a matrix and metadata
clustify(
    input = pbmc_matrix_small,
    metadata = pbmc_meta,
    ref_mat = cbmc_ref,
    query_genes = pbmc_vargenes,
    cluster_col = "classified",
    verbose = TRUE
)

# Annotate using a different method
clustify(
    input = pbmc_matrix_small,
    metadata = pbmc_meta,
    ref_mat = cbmc_ref,
    query_genes = pbmc_vargenes,
    cluster_col = "classified",
    compute_method = "cosine"
)

# Annotate a Seurat object
clustify(
    s_small3,
    cbmc_ref,
    cluster_col = "RNA_snn_res.1",
    obj_out = TRUE,
    per_cell = FALSE,
    dr = "tsne"
)

# Annotate (and return) a Seurat object per-cell
clustify(
    input = s_small3,
    ref_mat = cbmc_ref,
    cluster_col = "RNA_snn_res.1",
    obj_out = TRUE,
    per_cell = TRUE,
    dr = "tsne"
)

NCBI-Hackathons/RClusterCT documentation built on July 20, 2021, 3:15 p.m.