Skip to contents

Create a muscomic object

Source: R/objects.R
CreateMuscomicObject.Rd

Create a muscomic object containing coverage and (optionally) allelic count information for a single-cell omic dataset.

Usage

CreateMuscomicObject(
  type = c("ATAC", "RNA"),
  mat_counts,
  features,
  allele_counts = NULL,
  label.omic = NULL,
  label.features = NULL
)

Arguments

type

Type of single cell omic, either "ATAC" or "RNA" (character string).

mat_counts

Matrix of raw counts cells x features (matrix or dgCMatrix). Rows are cells and columns are features. (they must correspond to the id column of features)

features

Data frame of features (peaks, genes...) coordinates on genome (data.frame). It should contain 4 columns:

CHROM

Chromosome names in character format, e.g. "15", "X" (character).

start

Start positions (integer).

end

End positions (integer).

id

Unique identifiers, e.g. gene name "CDH1" or peak identifier CHROM_start_end "1_1600338_1600838" (character). It should match the feature identifiers as row names of mat_counts.

allele_counts

Data frame of allele counts at variant positions per cell (data.frame). Variant positions can be either common single nucleotide polymorphisms (SNPs) positions or individual-specific heterozygous positions retrieved from bulk sequencing. The data frame format is based on the Variant Calling Format (VCF), thereby it must contain the following columns : cell, CHROM, POS, REF, ALT, RD, AD. See exdata_allele_counts() for details.

label.omic

Label for the single cell omic (character string). By default "scATAC-seq" is used for "ATAC" type and "scRNA-seq" for "RNA" type.

label.features

Label for features (character string). By default "peaks" is used for "ATAC" type and "genes" for "RNA" type.

Value

A muscomic object.

See also

Examples


# Minimal example
mat <- Matrix::sparseMatrix(
    i = c(1, 1, 2),
    j = c(1, 2, 1),
    x = c(5, 3, 2),
    dims = c(2, 2),
    dimnames = list(c("cell1", "cell2"), c("f1", "f2"))
)

features <- data.frame(
    CHROM = c("1", "1"),
    start = c(100, 200),
    end   = c(150, 250),
    id    = c("f1", "f2")
)

allele <- data.frame(
    cell = c("cell1", "cell2"),
    CHROM = c("1", "1"),
    POS = c(100, 200),
    REF = c("A", "G"),
    ALT = c("C", "T"),
    RD = c(10, 5),
    AD = c(3, 1)
)

muscomic <- CreateMuscomicObject(
    type = "ATAC",
    mat_counts = mat,
    features = features,
    allele_counts = allele
)

# On the example dataset
atac <- CreateMuscomicObject(
  type = "ATAC",
  mat_counts = exdata_mat_counts_atac_tumor,
  allele_counts = exdata_allele_counts_atac_tumor,
  features = exdata_peaks
)
atac
#> A muscomic object 
#>  type: ATAC 
#>  label: scATAC-seq 
#>  cells: 71 
#>  counts: 71 cells x 1200 features (peaks)
#>  logratio: None
#>  variant positions: 681

# without allele counts data (not required for clustering step)
atac2 <- CreateMuscomicObject(
  type = "ATAC",
  mat_counts = exdata_mat_counts_atac_tumor,
  features = exdata_peaks
)
atac2
#> A muscomic object 
#>  type: ATAC 
#>  label: scATAC-seq 
#>  cells: 71 
#>  counts: 71 cells x 1200 features (peaks)
#>  logratio: None
#>  variant positions: 0