R/mirMrnaInt.R

In Krutik6/TimiRGeN: Time sensitive microRNA-mRNA integration, analysis and network generation tool

#' @title mirMrnaInt
#' @description Create a correlation matrix of all the potential miR-mRNA
#' interactions which could arise between the input miR data and the mRNAs
#' found from the wikiMrna function. The time series DE data will be averaged
#' from the dataframe created by diffExpressRes of miR data and the dataframe
#' created by wikiMrna. This will show miR-mRNA correlations over the time
#' course.
#' @param MAE MultiAssayExperiment which will store the output of mirMrnaInt.
#' It is recommended to begin a new MAE using MultiAssayExperiment() here so
#' the MAE objects do not get too large.
#' @param miR_express Dataframe from using the diffExpressRes function on miR
#' data. Rownames should be miR gene names and columns should include DE results
#' displaying abundance e.g. log2fc or average expression. These dataframes
#' should also have gene IDs. This should be stored as an assay within the
#' MAE used in the diffExpressRes function.
#' @param GenesofInterest Dataframe including mRNAs found in both the input data
#' and the pathway of interest, as well as gene IDs. This is the output from
#' wikiMrna. This should be found as an assay within the MAE which was used in
#' the wikiMrna function. Make sure the same ID type is used in the inputs
#' for miR_express and GenesofInterest.
#' @param maxInt Integer. Should be equal to number of samples in both mRNA and
#' miR data e.g. number of different time points. In the example it is 5 because
#' there are 5 time points.
#' @param corMeth Add string : "pearson", "spearman" or "kendall". Default is
#' "pearson".
#' @return A large correlation matrix which contains averaged miR-mRNA
#' time series information for every possible miR-mRNA interaction between the
#' genes of interest and all the miRs. Output will be stored as an assay in
#' the input MAE.
#' @export
#' @usage mirMrnaInt(MAE, miR_express, GenesofInterest, maxInt, corMeth)
#' @examples
#' G <- data.frame(row.names = c("Acaa1a", "Acadm", "Acss1", "Adh1"),
#'                 "D1.Log2FC" = c("-1.2944593","-2.0267432","-2.1934942",
#'                                 "-2.1095853"),
#'                 "D2.Log2FC" = c("-1.1962396","-2.1345451","-1.7699232",
#'                                 "-1.0961674"),
#'                 "D3.Log2FC" = c("0.2738496","-1.9991046","-1.7637549",
#'                                 "-1.6572653"),
#'                 "D7.Log2FC" = c("-0.51765245","-2.20689661","-0.68479699",
#'                                 "-2.06512466"),
#'                 "D14.Log2FC" = c("-0.4510294","-1.1523849","-0.4297012",
#'                                  "-1.1017597"),
#'                 "ID" = c("113868","11364","68738","11522"))
#'
#' MIR <- data.frame(row.names = c("mmu-miR-101a-3p", "mmu-miR-101a-5p",
#'                                 "mmu-miR-101c", "mmu-miR-106a-5p"),
#'                   "D1.Log2FC" = c("-0.0039141722","-0.4328659746",
#'                                   "-0.0038897133", "-0.4161749123"),
#'                   "D2.Log2FC" = c("-0.210605345","-0.600422732",
#'                                   "-0.210574742", "-0.530311376"),
#'                   "D3.Log2FC" = c("-0.315070839","-0.745367163",
#'                                   "-0.315012148", "-0.559274530"),
#'                   "D5.Log2FC" = c("-0.41087763","-0.63952382",
#'                                   "-0.41087876", "-1.03618015"),
#'                   "D14.Log2FC" = c("-0.39466968","-0.60122678",
#'                                    "-0.39461099", "-0.41889698"),
#'                   "ID" = c("387143","387143","100628572","723829"))
#'
#'MAE <- MultiAssayExperiment()
#'
#'MAE <- mirMrnaInt(MAE, miR_express = MIR, GenesofInterest = G,
#'                      maxInt = 5, corMeth = "pearson")
mirMrnaInt <- function(MAE, miR_express, GenesofInterest, maxInt, corMeth = "pearson"){


        if (missing(MAE)) stop('MAE is missing. Add MAE. Output from mirMrnaInt will be stored in the MAE. Please used the diffExpressRes and wikiMrna functions first.')

        if (missing(miR_express)) stop("miR_express is missing. Add a dataframe with miR abundance values and gene IDs. Please use the diffExpressRes function on miR data first. Output of diffExpressRes will be stored as an assay within the MAE used in the diffExpressRes function.")

        if (missing(GenesofInterest)) stop("GenesofInterest is missing. Add a dataframe containing mRNAs found in the wikipathway of interest and in the input mRNA data. Please use the wikiMrna function. Results from wikiMrna will be stored as an assay the MAE used in the wikiMrna function.")

        if (missing(maxInt)) stop("maxInt is missing. Add an integer which represents the number of time points / samples in your data set.")

        #Make matrix of the names of miRs and mRNAs of interest
        allnames <- expand.grid(rownames(miR_express),
                                rownames(GenesofInterest),
                                stringsAsFactors = FALSE)

        # Apply CorrTable to this
        interactions_df <- do.call(rbind, apply(allnames, 1,
                                                corrTable,
                                                miR_express,
                                                GenesofInterest,
                                                maxInt = maxInt,
                                                corMeth = corMeth))

        MAE2 <- suppressMessages(MultiAssayExperiment(list(
                                      interactions_df = interactions_df)))

        MAE <- c(MAE, MAE2)
return(MAE)
}