Add files via upload

DESCRIPTION

@@ -0,0 +1,23 @@
+Package: NOREVA
+Type: Package
+Title: R Package for Systematic Optimization of Metabolomic Data
+        Processing
+Version: 2.1.1
+Author: Jianbo Fu
+Maintainer: Jianbo Fu <[email protected]>
+Description: The NOREVA package not only enables the pre-processing and assessment of multi-class/time-series metabolomic data but also realize a high-throughput discovery of the well-performing pre-processing workflows.
+Depends: R (>= 3.5.0), rJava (>= 0.5-0)
+License: GPL
+Encoding: UTF-8
+LazyData: true
+RoxygenNote: 7.1.1
+Imports: DiffCorr, affy, vsn, DT, ropls, e1071, AUC, impute, eulerr,
+        MetNorm, ggsci, timecourse, multiROC, dummies, ggplot2, ggord,
+        limma, ggfortify, usethis, ggrepel, ggpubr, sampling, crmn,
+        ProteoMM, dplyr, statTarget, NormalizeMets, RcmdrMisc,
+        reshape2, foreach, data.table, parallel, doSNOW, tidyverse,
+        iterators
+Suggests: knitr, rmarkdown, testthat
+VignetteBuilder: knitr
+NeedsCompilation: yes
+Packaged: 2021-10-04 15:23:19 UTC; moumj

NAMESPACE

@@ -0,0 +1,110 @@
+# Generated by roxygen2: do not edit by hand
+
+export(EIGENMS)
+export(PrepareInuputFiles)
+export(loplot)
+export(normulticlassisall)
+export(normulticlassisallgs)
+export(normulticlassispart)
+export(normulticlassmatrix)
+export(normulticlassnoall)
+export(normulticlassnoallgs)
+export(normulticlassnopart)
+export(normulticlassqcall)
+export(normulticlassqcallgs)
+export(normulticlassqcpart)
+export(nortimecourseisall)
+export(nortimecourseisallgs)
+export(nortimecourseispart)
+export(nortimecoursematrix)
+export(nortimecoursenoall)
+export(nortimecoursenoallgs)
+export(nortimecoursenopart)
+export(nortimecourseqcall)
+export(nortimecourseqcallgs)
+export(nortimecourseqcpart)
+export(norvisualization)
+export(shiftCor)
+import(AUC)
+import(DT)
+import(DiffCorr)
+import(MetNorm)
+import(affy)
+import(crmn)
+import(doSNOW)
+import(dplyr)
+import(dummies)
+import(e1071)
+import(foreach)
+import(ggfortify)
+import(ggord)
+import(ggplot2)
+import(ggpubr)
+import(ggrepel)
+import(ggsci)
+import(impute)
+import(iterators)
+import(limma, except=.__C__LargeDataObject)
+import(multiROC)
+import(parallel)
+import(rJava)
+import(ropls, except=plot)
+import(sampling)
+import(statTarget)
+import(timecourse)
+import(usethis)
+import(vsn)
+importFrom(NormalizeMets,LinearModelFit)
+importFrom(ProteoMM,eig_norm1)
+importFrom(ProteoMM,eig_norm2)
+importFrom(RcmdrMisc,numSummary)
+importFrom(data.table,data.table)
+importFrom(grDevices,colorRampPalette)
+importFrom(grDevices,dev.off)
+importFrom(grDevices,pdf)
+importFrom(grDevices,png)
+importFrom(grDevices,rainbow)
+importFrom(grDevices,rgb)
+importFrom(graphics,abline)
+importFrom(graphics,close.screen)
+importFrom(graphics,legend)
+importFrom(graphics,lines)
+importFrom(graphics,mtext)
+importFrom(graphics,par)
+importFrom(graphics,plot)
+importFrom(graphics,points)
+importFrom(graphics,screen)
+importFrom(graphics,split.screen)
+importFrom(graphics,symbols)
+importFrom(graphics,text)
+importFrom(graphics,title)
+importFrom(reshape2,melt)
+importFrom(ropls,opls)
+importFrom(stats,anova)
+importFrom(stats,as.formula)
+importFrom(stats,cor)
+importFrom(stats,dnorm)
+importFrom(stats,kmeans)
+importFrom(stats,lm)
+importFrom(stats,loess)
+importFrom(stats,loess.control)
+importFrom(stats,mad)
+importFrom(stats,median)
+importFrom(stats,model.matrix)
+importFrom(stats,na.omit)
+importFrom(stats,pf)
+importFrom(stats,pnorm)
+importFrom(stats,qnorm)
+importFrom(stats,qt)
+importFrom(stats,quantile)
+importFrom(stats,resid)
+importFrom(stats,rnorm)
+importFrom(stats,runif)
+importFrom(stats,sd)
+importFrom(stats,var)
+importFrom(utils,combn)
+importFrom(utils,read.csv)
+importFrom(utils,setTxtProgressBar)
+importFrom(utils,txtProgressBar)
+importFrom(utils,write.csv)
+importFrom(utils,write.table)

R/CS_multi_class.R

@@ -0,0 +1,19 @@
+#' @importFrom ropls opls
+options(warn = -1)
+OPLSDA_C <- function(mat, lab) {
+  X <- mat
+  Y <- as.factor(lab)
+
+  # !!! adding a FALSE for scale.
+  # For the following PLS-DA model, the samples in row and variables in column.
+  #oplsda <- opls(X, Y, permI = 100, predI = 2, scaleC = "standard", crossvalI=2, plot = FALSE)
+  oplsda <- ropls::opls(X, Y, permI = 100, predI = 2, scaleC = "standard", crossvalI=2, fig.pdfC = FALSE)
+  #res <- oplsda$vipVn
+  res <- oplsda@vipVn
+  cpds <- data.frame(CompoundName = names(res), VIP = res)
+
+  cpds_filter <- cpds[order(as.numeric(cpds[,2]), decreasing=T),1]
+
+  return(cpds_filter)
+}
+##########################################################

R/EIGENMS.R

@@ -0,0 +1,53 @@
+#' @title EigenMS normalization
+#' @description EigenMS is an adaptation of surrogate variable analysis, which identifies trends
+#' attributable to bias by utilizing singular value decomposition on model residuals.
+#' See the details at the following References.
+#' @param data Input matrix of data
+#' @param label Input the label of data
+#' @return A structure with multiple components
+#' \describe{
+#'   \item{m_logInts}{number of metabolites x number of samples
+#'            matrix of expression data with no missing values}
+#'   \item{grps}{either a single factor indicating the treatment group of
+#'           each sample i.e. [1 1 1 1 2 2 2 2...]
+#'           or a data frame of factors}
+#'   \item{m_ints_eig1}{First portion of EigenMS: Identify eigentrends attributable to bias}
+#'   \item{m_ints_norm1}{Eliminate the effects of systematic bias identified in eig_norm1()}
+#'   \item{mm_eigenMS}{matrix of normalized abundances, no extra columns}
+#'}
+#' @importFrom ProteoMM eig_norm1
+#' @importFrom ProteoMM eig_norm2
+#' @usage EIGENMS(data,label)
+#' @examples
+#' \donttest{data(mm_metabolites)}
+#' \donttest{head(mm_metabolites)}
+#' # different from parameter names as R uses outer name spaces
+#' # if variable is undefined
+#' \donttest{intsCols = 8:13}
+#' \donttest{metaCols = 1:7} # reusing this variable
+#' \donttest{m_logInts = make_intencities(mm_metabolites, intsCols)}  # will reuse the name
+#' \donttest{m_logInts = convert_log2(m_logInts)}
+#' # 3 samples for CG and 3 for mCG
+#' \donttest{grps = as.factor(c('CG','CG','CG', 'mCG','mCG','mCG'))}
+#' \donttest{mm_eigenMS = EIGENMS(m_logInts,grps)}
+#' @references 1. Metabolomics data normalization with EigenMS.
+#' Karpievitch YK, Nikolic SB, Wilson R, Sharman JE, Edwards LM. 2014, PLoS ONE.
+#' @references 2. Normalization of peak intensities in bottom-up MS-based proteomics
+#' using singular value decomposition.
+#' Karpievitch YV, Taverner T et al. 2009, Bioinformatics.
+
+#' @export
+
+EIGENMS <- function(data, label) {
+
+  a <- data
+  ddata<-a
+  m_logInts = ddata
+  grps = as.factor(label)
+  m_prot.info = cbind(rownames(ddata),rownames(ddata))
+  m_ints_eig1 = suppressWarnings(suppressMessages(eig_norm1(m=m_logInts,treatment=grps,prot.info=m_prot.info)))
+  m_ints_norm1 = suppressWarnings(suppressMessages(eig_norm2(rv=m_ints_eig1)))
+  eigenMS <-m_ints_norm1$norm_m
+  return(eigenMS)
+
+}

R/Evaluation.R

@@ -0,0 +1,173 @@
+
+###==========================================================================###
+### Other evaluation indexes, such as PEV, PMAD, PCV, and so on.
+###==========================================================================###
+
+#calculate PEV
+PEV <- function(data0){
+    data0<-as.matrix(data0)
+    data0<-data0[order(data0[,1]),]
+    label <- as.factor(data0[,1])
+    data <- data0[,-1]
+    data <- t(as.matrix(data))
+    x<-levels(label)[1]
+    z<-1
+    y<-1
+    flag<-1
+    count<-0
+    varmem<-vector()
+    tempvar<-vector()
+    nonmissingmat<-vector()
+    for(i in 1:length(label))
+    {
+        if(x!=label[i] || i==length(label))
+        {
+            y<-i-1
+            if(i==length(label))
+            {
+                y<-i
+            }
+            if(flag==1)
+            {
+                count<-count+1
+                nonmissingmat<-(apply(data[,z:y],1,function(x) {((sum(!is.na(x))))}))-1
+                tempvar<-nonmissingmat*apply(data[,z:y],1,function(x) {var(x,na.rm=TRUE)})
+            }
+            if(flag==2)
+            {
+                count<-count+1
+                nonmissingmat<-(apply(data[,z:y],1,function(x) {((sum(!is.na(x))))}))-1
+                tempvar<-nonmissingmat*apply(data[,z:y],1,function(x) {var(x,na.rm=TRUE)})
+            }
+            varmem<-c(varmem,((sum(tempvar,na.rm=T))/(sum(nonmissingmat,na.rm=T))))
+            z<-i
+            x<-label[i]
+            flag=2;
+        }
+    }
+    avgvarmem<-varmem
+    names(avgvarmem)<-levels(label)
+    return(avgvarmem)
+}
+#calculate PMAD
+PMAD <- function(data0){
+    data0<-as.matrix(data0)
+    data0<-data0[order(data0[,1]),]
+    label <- as.factor(data0[,1])
+    data <- data0[,-1]
+    data <- t(as.matrix(data))
+    data <- apply(data, 1:2, as.numeric)
+    x<-levels(label)[1]
+    z<-1
+    y<-1
+    flag<-1
+    count<-0
+    madmem<-matrix(nrow=nrow(data),ncol=length(levels(as.factor(unlist(label)))),byrow=T)
+    for(i in 1:length(label))
+    {
+        if(x!=label[i] || i==length(label))
+        {
+            y<-i-1
+            if(i==length(label))
+            {
+                y<-i
+            }
+            if(flag==1)
+            {
+                count<-count+1
+                madmem[,count]<-apply(data[,z:y],1,function(x) {mad(x,na.rm=T)})
+            }
+            if(flag==2)
+            {
+                count<-count+1
+                madmem[,count]<-apply(data[,z:y],1,function(x) {mad(x,na.rm=T)})
+            }
+            z<-i
+            x<-label[i]
+            flag=2;
+        }
+    }
+    avgmadmem<-apply(madmem,2,mean,na.rm=T)
+    names(avgmadmem)<-levels(label)
+    return(avgmadmem)
+}
+
+#' @importFrom RcmdrMisc numSummary
+#calculate PCV
+PCV <- function(data0){
+    data0<-as.matrix(data0)
+    data0<-data0[order(data0[,1]),]
+    label <- as.factor(data0[,1])
+    data <- data0[,-1]
+    data <- t(as.matrix(data))
+    data <- apply(data, 1:2, as.numeric)
+
+    tempcvmat<-matrix(nrow=nrow(data),ncol=length(levels(as.factor(unlist(label)))),byrow=T)
+    for(i in 1:nrow(data))
+    {
+        tempcv<-numSummary(data[i,],statistics=c("cv"),groups=unlist(label))
+        tempcvmat[i,]<-tempcv$table
+    }
+    temcvmatsum<-apply(tempcvmat,2,mean,na.rm=T)
+    avgcvmem <-((temcvmatsum*100))
+    names(avgcvmem)<-levels(label)
+    return(avgcvmem)
+}
+
+
+
+  purity<-function(result,label){
+    total_num<-length(label)
+    cluster_counter = unique(result)
+    original_counter=unique(label)
+    t<-NULL
+
+    for (k in cluster_counter){
+      p_k<-NULL
+      for (j in original_counter){
+        count<-0
+        for (i in 1:length(result)){
+          if (result[i]==k && label[i]==j){
+            count<-count+1
+          }
+
+        }
+        p_k<-c(p_k,count)
+      }
+      temp_t<-max(p_k)
+      t<-c(t,temp_t)
+    }
+    res<-sum(t)/total_num
+    return(res)
+
+  }
+
+
+
+  CWvalue <- function(all_genelist=all_genelist,Y,n){
+    num <- 3
+    partial_genelist <- vector("list", length(all_genelist))
+    for (j in 1:length(all_genelist)) {
+      partial_genelist[[j]] <- all_genelist[[j]][1:n]
+    }
+
+    partial_genelist <- as.list(partial_genelist)
+    partial_genelist <- table(unlist(partial_genelist))
+
+    Ff_sum <- 0
+    for (k in 1:length(partial_genelist)) {
+      Ff_sum <- Ff_sum + partial_genelist[k] * (partial_genelist[k]-1)
+    }
+    Ff_sum <- as.numeric(Ff_sum)
+    #Y <- ncol(data)
+    N <- sum(partial_genelist)
+    D <- N %% Y
+    H <- N %% num
+
+    CWrel <- (Y*(N-D+Ff_sum)-N^2+D^2)/(Y*(H^2+num*(N-H)-D)-N^2+D^2)
+    return(CWrel)
+  }
+
+### End
+### ------------------------------------------------------------------------ ###
+