getMarker: Marker selection

In Fang0828/SCMarker: What the package does (short line)

Description

Selecting markers if gene pairs are mutual nearest coexpressed or exclusivity in a number of cells.

Usage

getMarker(obj, k=300, n=30)

Arguments

obj	The list oject from function GeneFilter
k	The maximum number that how many nearest neighbors would be considered, default = 300.
n	The minimum number which is the coexpressed/mutually exclusivie cells for each gene pairs, default = 30

Value

dataThe raw expresion data after filter widly expressed genes geneSummA data frame including gene names as, corresponding number of expressed cells and the peak number of probability density distribution function cellSummA data frame including cell ID and corresponding number of expressed genes binadataThe binary expression matrix corresponding to raw data markerThe filnal marker selected

Examples

data(melanoma)
melanoma1=as.matrix(melanoma[,2:dim(melanoma)[2]])
row.names(melanoma1)=melanoma[,1]
res=ModalFilter(data=melanoma1,geneK=10,cellK=10,width=2,cutoff=2)
res=GeneFilter(obj=res)
res=getMarker(obj=res)
res=getMarker(obj=res,k=200,n=20)


## The function is currently defined as
function (obj, k=300, n=30)
{
    data = obj$newdata
    binadata = obj$binadata
    genename = row.names(binadata)
    geneindex = rowSums(binadata)
    HamD = tcrossprod(binadata)
    diag(HamD) = 0
    HamDD = tcrossprod((1 - binadata), binadata)
    MNNmarker = getMNN(HamD = HamD, genename = genename, k = k,
        n = n)
    MENmarker = getMEN(HamD = HamDD, genename = genename, k = k,
        n = n)
    obj$marker = union(MNNmarker, MENmarker)
    return(obj)
  }

Fang0828/SCMarker documentation built on May 13, 2019, 12:51 p.m.