Find top features and metabolites that are strongly correlated with a given feature

This function returns a list of features ranked by highest Pearson correlation score. The specified feature to corrleate against can be either a m/z value, gene or an ident (i.e. cluster). For multi-omic data, both a metabolic and transcriptomic assay can be specified to calculate correlation of both metabolites and genes.

Usage

FindCorrelatedFeatures(
  data,
  mz = NULL,
  gene = NULL,
  ident = NULL,
  SM.assay = "SPM",
  ST.assay = NULL,
  SM.slot = "counts",
  ST.slot = "counts",
  nfeatures = 10
)

Arguments

data: SpaMTP Seurat class object containing both Spatial Transcriptomic and Metabolic data assays.
mz: Numeric string specifying the m/z to find correlated features for. One of mz, gene or ident must be provided, alternatives must be NULL (default = NULL).
gene: Character string specifying the gene to find correlated features for. One of mz, gene or ident must be provided, alternatives must be NULL (default = NULL).
ident: Character string defining the ident column in the data object's @meta.data slot to find correlated features for. One of mz, gene or ident must be provided, alternatives must be NULL (default = NULL).
SM.assay: Character string specifying the name of the assay containing the spatial metabolomics (SM) data (default = "SPM").
ST.assay: Character string specifying the name of the assay containing the spatial transcriptomics (ST) data. If NULL then only metabolites will be used (Default = NULL).
SM.slot: Character string specifying the slot of the SM assay to use (default = "counts").
ST.slot: Character string specifying the slot of the ST assay to use (default = "counts").
nfeatures: Integer specifying the number of top correlated features to return (default = 10).

Value

A data frame containing the top correlated features with columns for the feature names and their correlation values.

Examples

# result <- FindCorrelatedFeatures(data = SpaMTP, gene = "GeneX", nfeatures = 5)