R/get_score.R

In tinglabs/reCAT: Reconstructing Cell Cycle Pseudo Time-Series via Single-Cell Transcriptome Data

#' Bayes scores and mean scores
#'
#' This function produces the bayes scores (G1, S, G2/M) and Mean scores (G1, S, G2, M, G1/S, G2/M) over the dataset. The output can then be used to classify different cycle stages using HMM.
#'@param data The data with cycling genes, m genes by n cells
#'@keywords get_test_exp
#'@export
#'@examples
#'test_exp <- get_test_exp(data)
#'get_score(t(test_exp))

# in this function, you should input a matrix that you want to calculate bayes_score and mean_score.

# the matrix is transposition of test_exp(you use in get the ordIndex)

# ola_mES_2i.RData in R is a sample


#source("bayes_score_function.R")
#source("mean_score_function.R")
#load("../data/tranning.RData")
#load("../data/ccgene.RData")

get_score <- function(talentDE)
{
    rname = row.names(talentDE)
    if ((length(grep("ensg", tolower(rname[1]))) == 0) & (length(grep("ensmusg", tolower(rname[1]))) == 0))
    {
        rname = geneReferTable[match(tolower(rname), tolower(geneReferTable[,4])),3]
    }else
    {
        if (length(grep("ensg", tolower(rname[1]))) != 0)
        {
            rname = geneReferTable[match(tolower(rname), tolower(geneReferTable[,1])),3]
        }
    }
    row.names(talentDE) = rname

    y <- rep(0,length(training.data[1,]))
    cname <- colnames(training.data)
    G1.id <- grep("G1", cname)
    S.id <- grep("S", cname)
    G2M.id <- grep("G2", cname)
    y[G1.id] <- 1
    y[S.id] <- 2
    y[G2M.id] <- 3

    markers <- getPairs(talentDE, G1.marker.pairs,
                        S.marker.pairs, G2M.marker.pairs, genes.training)
    vector.training <- t(getVector(training.data, markers))

    vector.data <- t(getVector(talentDE, markers))

    pred <- Naive(vector.training, length(G1.id), length(S.id), length(G2M.id), dim(training.data)[2], vector.data)

    result <- averageScoreall(talentDE, allgene)

    return(list(bayes_score = pred, mean_score = result))
}