Skip to contents

To build an metacell object from the given metacell partitioning

mcRigor_buildmc.Rd

To build an metacell object from the given metacell partitioning (sc_membership)

Usage

mcRigor_buildmc(
  obj_singlecell,
  sc_membership = NULL,
  assay_type = c("RNA", "ATAC"),
  doNorm = T,
  aggregate_method = c("mean", "sum", "geom"),
  doAssign = T,
  fields = NULL,
  covariate_method = "absolute",
  purity_method = "max_proportion",
  add_testres = F,
  test_stats = NULL,
  Thre = NULL,
  test_cutoff = 0.01,
  prePro = F,
  feature_use = 2000,
  gene_filter = 0.1,
  cor_method = c("pearson", "spearman")
)

Arguments

obj_singlecell

the Seurat object of single cells

sc_membership

A named vector (or dataframe) of the metacell membership of single cells

assay_type

The type of data assay yuo are using, depending on which different normalization would be used.

doNorm

A bool indicating whether to perform normalization for the metacell object (obj_metacell)

aggregate_method

The method to aggregate single cell profiles into metacell profiles

doAssign

A bool indicating whether to assign covariates to metacells or not

fields

A vector of covariate names to assign

covariate_method

Method to define the most abundant cell covariate within metacells. Available: "jaccard", "relative", "absolute" (default).

  • jaccard - assign metacell to covariate with the maximum jaccard coefficient (recommended)

  • relative - assign metacell to covariate with the maximum relative abundance (normalized by cluster size), may result in assignment of metacells to poorly represented (small) covariate due to normalization

  • absolute - assign metacell to covariate with the maximum absolute abundance within metacell, may result in disappearance of poorly represented (small) clusters

purity_method

method to compute metacell purity. "max_proportion" if the purity is defined as a proportion of the most abundant covariate (cell type) within super-cell or "entropy" if the purity is defined as the Shanon entropy of the covariates metacell consists of.

add_testres

A bool indicating whether to add the mcRigor results (dubious or trustworthy) as part of obj_metacell's metadata

test_stats

If add_testres = True, this argument is needed. Usually should be TabMC, an output from previous steps.

Thre

The threshold for dubious metacell detection. If not inputed, it will be computed based on test_stats.

test_cutoff

The test size for dubious metacell detection testing

prePro

A boolean indicating whether to normalize obj_singlecell for preprocessing.

feature_use

The number of genes to use in metacell testing.

gene_filter

A proportion. Genes expressed lower than this proportion will be filtered out.

cor_method

The method for gene correlation calculation description

Value

a Seurat object of the metacells