OmicsProfiles

Profile(countmat, omicsname, var, obs; varindex=:genesymbol, obsindex=:barcode,
        T=float(eltype(countmat)))

Constructor for establishing AnnotatedProfile with a OmicsProfile inside.

Arguments

• countmat: The count matrix for given omics and it will be set to key :count.
• omicsname::Symbol: The name of given OmicsProfile.
• var::DataFrame: The dataframe contains meta information for features or variables.
• obs::DataFrame: The dataframe contains meta information for observations.
• varindex::Symbol: The index of dataframe var.
• obsindex::Symbol: The index of dataframe obs.
• T: Element type of countmat.

Examples

julia> ngenes, ncells = (100, 500)
(100, 500)

julia> data = rand(0:10, ngenes, ncells);

julia> var = DataFrame(genesymbol=1:ngenes);

julia> obs = DataFrame(barcode=1:ncells);

julia> prof = Profile(data, :RNA, var, obs, varindex=:genesymbol, obsindex=:barcode)
AnnotatedProfile (nobs = 500):
    obs: barcode
RNA => OmicsProfile (nvar = 100):
    var: genesymbol
    layers: count

See also AnnotatedProfile for multi-omics data container and OmicsProfile for single omics data container.

OmicsProfile(countmat, var, varindex; T=float(eltype(countmat)))

A data container preserves single omics data for single cell sequencing analysis. Read count data countmat and variables/features metadata var are retained in the container. It records data and progression mainly about variables or features.

Arguments

• countmat: The count matrix for given omics and it will be set to key :count.
• var::DataFrame: The dataframe contains meta information for features or variables.
• varindex::Symbol: The index of dataframe var.
• T: Element type of countmat.

Examples

julia> ngenes, ncells = (100, 500)
(100, 500)

julia> data = rand(0:10, ngenes, ncells);

julia> var = DataFrame(index=1:ngenes, C=rand(ngenes), D=rand(ngenes));

julia> prof = OmicsProfile(data, var, :index)
OmicsProfile (nvar = 100):
    var: index, C, D
    layers: count

See also AnnotatedProfile for multi-omics data container.

AnnotatedProfile(omics, obs, obsindex)
AnnotatedProfile(op, omicsname, obs, obsindex)

A data container preserves multi-omics data for single cell sequencing analysis. Multi-omics omics in dictionary or, in separation, single omics op with its name omicsname and observations metadata obs are retained in the container. It records data and progression mainly about observations.

Arguments

• omics::Dict{Symbol,OmicsProfile}: A collection of omics profile with their names in keys.
• op::OmicsProfile: Single omics profile.
• omicsname::Symbol: The name of given OmicsProfile.
• obs::DataFrame: The dataframe contains meta information for observations.
• obsindex::Symbol: The index of dataframe obs.

Examples

julia> ngenes, ncells = (100, 500)
(100, 500)

julia> data = rand(0:10, ngenes, ncells);

julia> var = DataFrame(genesymbol=1:ngenes);

julia> obs = DataFrame(barcode=1:ncells);

julia> omic = OmicsProfile(data, var, :genesymbol);

julia> ap = AnnotatedProfile(omic, :RNA, obs, :barcode)
AnnotatedProfile (nobs = 500):
    obs: barcode
RNA => OmicsProfile (nvar = 100):
    var: genesymbol
    layers: count

See also Profile for routine use and OmicsProfile for single omics data container.

read_10x(path; make_unique=true, omicsname=:RNA, varnames=[:ensembleid, :genesymbol],
    obsnames=[:barcode], varindex=:genesymbol, obsindex=:barcode)

Reads 10x dataset from path and returns an AnnotatedProfile.

Arguments

• path::AbstractString: The path to 10x dataset folder.
• make_unique::Bool: To make varindex unique or not if given varindex is not unique.
• omicsname::Symbol: The name of given sequencing data.
• varnames::AbstractVector: The header or column names of feature file.
• obsnames::AbstractVector: The header or column names of barcode file.
• varindex::Symbol: The index of dataframe var.
• obsindex::Symbol: The index of dataframe obs.

Examples

julia> using SnowyOwl

julia> prof = read_10x(SnowyOwl.Dataset.pbmc3k_folder(), varindex=:ensembleid)
AnnotatedProfile (nobs = 2700):
    obs: barcode
RNA => OmicsProfile (nvar = 32738):
    var: ensembleid, genesymbol
    layers: count