Merge pull request #9 from large-scale-gxe-methods/dev

Dev

DESCRIPTION

@@ -12,7 +12,7 @@ Description: Use a 'glmmkin' class object (GMMAT package) from the null model to
 License: GPL-3
 Copyright: See COPYRIGHTS for details.
 Imports: Rcpp, Matrix, parallel, MASS, SeqArray, SeqVarTools, foreach,
-        GMMAT, CompQuadForm
+        GMMAT, CompQuadForm, data.table
 Suggests: doMC, testthat
 LinkingTo: Rcpp, RcppArmadillo
 Encoding: UTF-8

NAMESPACE

@@ -7,6 +7,7 @@ importFrom("stats", "as.formula", "binomial", "dbeta", "glm", "model.frame",
 "pnorm", "uniroot", "integrate", "weights", "vcov")
 importFrom("utils", "read.table", "write.table")
 importFrom("CompQuadForm", "davies", "liu")
+importFrom("data.table", "fread")
 importFrom("SeqArray", "seqOpen", "seqGetData", "seqClose", "seqSetFilter")
 importFrom("SeqVarTools", "missingGenotypeRate", "alleleFrequency", "altDosage")
 importFrom("GMMAT", "glmmkin", "glmm.score", "SMMAT")

R/MAGEE.R

@@ -1,4 +1,4 @@
-MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep = "\t", bgen.samplefile = NULL, interaction.covariates = NULL, meta.file.prefix = NULL, MAF.range = c(1e-7, 0.5), MAF.weights.beta = c(1, 25), miss.cutoff = 1, missing.method = "impute2mean", method = "davies", tests = "JF", use.minor.allele = FALSE, auto.flip = FALSE, Garbage.Collection = FALSE, is.dosage = FALSE, ncores = 1){
+MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep = "\t", bgen.samplefile = NULL, interaction.covariates = NULL, meta.file.prefix = NULL, MAF.range = c(1e-7, 0.5), AF.strata.range = c(0, 1), MAF.weights.beta = c(1, 25), miss.cutoff = 1, missing.method = "impute2mean", method = "davies", tests = "JF", use.minor.allele = FALSE, auto.flip = FALSE, Garbage.Collection = FALSE, is.dosage = FALSE, ncores = 1){
   if(Sys.info()["sysname"] == "Windows" && ncores > 1) {
     warning("The package doMC is not available on Windows... Switching to single thread...")
     ncores <- 1
@@ -86,7 +86,7 @@ MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep =
     }
     variant.id <- paste(chr, pos, ref, alt, sep = ":")
     rm(chr, pos, ref, alt); gc()
-    group.info <- try(read.table(group.file, header = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
+    group.info <- try(fread(group.file, header = FALSE, data.table = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
     if (inherits(group.info, "try-error")) {
       stop("Error: cannot read group.file!")
     }
@@ -176,7 +176,7 @@ MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep =
           if(!is.null(strata)) { # E is not continuous
             freq.tmp <- sapply(strata.list, function(x) colMeans(geno[x, , drop = FALSE], na.rm = TRUE)/2) # freq.tmp is a matrix, each column is a strata, and each row is a varirant 
             if (length(dim(freq.tmp)) == 2) freq_strata <- apply(freq.tmp, 1, range) else freq_strata <- as.matrix(range(freq.tmp)) # freq_strata is the range of allele freq across strata.list
-            include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= MAF.range[1] & freq_strata[2,] <= 1-MAF.range[1]
+            include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= AF.strata.range[1] & freq_strata[2,] <= AF.strata.range[2]
             rm(freq.tmp)
           }
           n.p <- sum(include)
@@ -385,7 +385,7 @@ MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep =
         if(!is.null(strata)) { # E is not continuous
           freq.tmp <- sapply(strata.list, function(x) colMeans(geno[x, , drop = FALSE], na.rm = TRUE)/2) # freq.tmp is a matrix, each column is a strata, and each row is a varirant 
           if (length(dim(freq.tmp)) == 2) freq_strata <- apply(freq.tmp, 1, range) else freq_strata <- as.matrix(range(freq.tmp)) # freq_strata is the range of allele freq across strata.list
-          include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= MAF.range[1] & freq_strata[2,] <= 1-MAF.range[1]
+          include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= AF.strata.range[1] & freq_strata[2,] <= AF.strata.range[2]
           rm(freq.tmp)
         }
         n.p <- sum(include)
@@ -551,7 +551,7 @@ MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep =
       if (is.null(bgen.samplefile)) {
         stop("Error: bgen file does not contain sample identifiers. A .sample file (bgen.samplefile) is needed.")
       }
-      sample.id <- read.table(bgen.samplefile, header = TRUE, sep = " ")
+      sample.id <- fread(bgen.samplefile, header = TRUE, data.table = FALSE)
       if ((nrow(sample.id)-1) != bgenInfo$N){
         stop(paste0("Error: Number of sample identifiers in BGEN sample file (", nrow(sample.id)-1, ") does not match number of samples in BGEN file (", bgenInfo$N,")."))
       }
@@ -594,7 +594,7 @@ MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep =
     variant.id <- paste(bgenVariant$VariantInfo$CHR, bgenVariant$VariantInfo$POS, bgenVariant$VariantInfo$A1, bgenVariant$VariantInfo$A2, sep = ":")
     gc()
     variant.idx <- 1:length(variant.id)
-    group.info <- try(read.table(group.file, header = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
+    group.info <- try(fread(group.file, header = FALSE, data.table = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
     if (inherits(group.info, "try-error")) {
       stop("Error: cannot read group.file!")
     }
@@ -683,7 +683,7 @@ MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep =
           if(!is.null(strata)) { # E is not continuous
             freq.tmp <- sapply(strata.list, function(x) colMeans(geno[x, , drop = FALSE], na.rm = TRUE)/2) # freq.tmp is a matrix, each column is a strata, and each row is a varirant 
             if (length(dim(freq.tmp)) == 2) freq_strata <- apply(freq.tmp, 1, range) else freq_strata <- as.matrix(range(freq.tmp)) # freq_strata is the range of allele freq across strata.list
-            include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= MAF.range[1] & freq_strata[2,] <= 1-MAF.range[1]
+            include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= AF.strata.range[1] & freq_strata[2,] <= AF.strata.range[2]
             rm(freq.tmp)
           }
           n.p <- sum(include)
@@ -886,7 +886,7 @@ MAGEE <- function(null.obj, interaction, geno.file, group.file, group.file.sep =
         if(!is.null(strata)) { # E is not continuous
           freq.tmp <- sapply(strata.list, function(x) colMeans(geno[x, , drop = FALSE], na.rm = TRUE)/2) # freq.tmp is a matrix, each column is a strata, and each row is a varirant 
           if (length(dim(freq.tmp)) == 2) freq_strata <- apply(freq.tmp, 1, range) else freq_strata <- as.matrix(range(freq.tmp)) # freq_strata is the range of allele freq across strata.list
-          include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= MAF.range[1] & freq_strata[2,] <= 1-MAF.range[1]
+          include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= AF.strata.range[1] & freq_strata[2,] <= AF.strata.range[2]
           rm(freq.tmp)
         }
         n.p <- sum(include)
@@ -1217,7 +1217,7 @@ MAGEE.prep <- function(null.obj, interaction, geno.file, group.file, interaction
   }
   variant.id <- paste(chr, pos, ref, alt, sep = ":")
   rm(chr, pos, ref, alt); gc()
-  group.info <- try(read.table(group.file, header = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
+  group.info <- try(fread(group.file, header = FALSE, data.table = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
   if (inherits(group.info, "try-error")) {
     stop("Error: cannot read group.file!")
   }
@@ -1258,7 +1258,7 @@ MAGEE.prep <- function(null.obj, interaction, geno.file, group.file, interaction
   return(out)
 }
 
-MAGEE.lowmem <- function(MAGEE.prep.obj, geno.file = NULL, meta.file.prefix = NULL, MAF.range = c(1e-7, 0.5), MAF.weights.beta = c(1, 25), miss.cutoff = 1, missing.method = "impute2mean", method = "davies", tests = "JF", use.minor.allele = FALSE, Garbage.Collection = FALSE, is.dosage = FALSE, ncores = 1)
+MAGEE.lowmem <- function(MAGEE.prep.obj, geno.file = NULL, meta.file.prefix = NULL, MAF.range = c(1e-7, 0.5), AF.strata.range = c(0, 1), MAF.weights.beta = c(1, 25), miss.cutoff = 1, missing.method = "impute2mean", method = "davies", tests = "JF", use.minor.allele = FALSE, Garbage.Collection = FALSE, is.dosage = FALSE, ncores = 1)
 {
   if(!inherits(MAGEE.prep.obj, "MAGEE.prep")) stop("Error: MAGEE.prep.obj must be a class MAGEE.prep object!")
   is.Windows <- Sys.info()["sysname"] == "Windows"
@@ -1354,7 +1354,7 @@ MAGEE.lowmem <- function(MAGEE.prep.obj, geno.file = NULL, meta.file.prefix = NU
 	if(!is.null(strata)) { # E is not continuous
           freq.tmp <- sapply(strata.list, function(x) colMeans(geno[x, , drop = FALSE], na.rm = TRUE)/2) # freq.tmp is a matrix, each column is a strata, and each row is a varirant 
           if (length(dim(freq.tmp)) == 2) freq_strata <- apply(freq.tmp, 1, range) else freq_strata <- as.matrix(range(freq.tmp)) # freq_strata is the range of allele freq across strata.list
-          include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= MAF.range[1] & freq_strata[2,] <= 1-MAF.range[1]
+          include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= AF.strata.range[1] & freq_strata[2,] <= AF.strata.range[2]
           rm(freq.tmp)
         }
         n.p <- sum(include)
@@ -1557,7 +1557,7 @@ MAGEE.lowmem <- function(MAGEE.prep.obj, geno.file = NULL, meta.file.prefix = NU
       if(!is.null(strata)) { # E is not continuous
         freq.tmp <- sapply(strata.list, function(x) colMeans(geno[x, , drop = FALSE], na.rm = TRUE)/2) # freq.tmp is a matrix, each column is a strata, and each row is a varirant 
         if (length(dim(freq.tmp)) == 2) freq_strata <- apply(freq.tmp, 1, range) else freq_strata <- as.matrix(range(freq.tmp)) # freq_strata is the range of allele freq across strata.list
-        include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= MAF.range[1] & freq_strata[2,] <= 1-MAF.range[1]
+        include <- include & !is.na(freq_strata[1,]) & !is.na(freq_strata[2,]) & freq_strata[1,] >= AF.strata.range[1] & freq_strata[2,] <= AF.strata.range[2]
         rm(freq.tmp)
       }
       n.p <- sum(include)

R/MAGEE.meta.R

@@ -16,7 +16,7 @@ MAGEE.meta <- function(meta.files.prefix, n.files = rep(1, length(meta.files.pre
   JV <- "JV" %in% tests
   JF <- "JF" %in% tests
   JD <- "JD" %in% tests
-  group.info <- try(read.table(group.file, header = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
+  group.info <- try(fread(group.file, header = FALSE, data.table = FALSE, col.names = c("group", "chr", "pos", "ref", "alt", "weight"), colClasses = c("character","character","integer","character","character","numeric"), sep = group.file.sep), silent = TRUE)
   if (inherits(group.info, "try-error")) {
     stop("Error: cannot read group.file!")
   }
@@ -32,7 +32,7 @@ MAGEE.meta <- function(meta.files.prefix, n.files = rep(1, length(meta.files.pre
   for(i in 1:n.cohort) { # Read the scores for each study from each core
     tmp.scores <- NULL
     for(j in 1:n.files[i]) { # n.files[i] is the number of cores for the i-th study
-      tmp <- try(read.table(paste0(meta.files.prefix[i], ".score.", j), header = TRUE, as.is = TRUE))
+      tmp <- try(fread(paste0(meta.files.prefix[i], ".score.", j), header = TRUE, data.table = FALSE))
       if (inherits(tmp,"try-error")) {
         stop(paste0("Error: cannot read ", meta.files.prefix[i], ".score.", j, "!"))
       }

R/glmm.gei.R

@@ -651,7 +651,7 @@ glmm.gei <- function(null.obj, interaction, geno.file, outfile, bgen.samplefile=
       if (is.null(bgen.samplefile)) {
         stop("Error: bgen file does not contain sample identifiers. A .sample file (bgen.samplefile) is needed.")
       }
-      sample.id <- read.table(bgen.samplefile, header = TRUE, sep = " ")
+      sample.id <- fread(bgen.samplefile, header = TRUE, data.table = FALSE)
       if ((nrow(sample.id)-1) != bgenInfo$N){
         stop(paste0("Error: Number of sample identifiers in BGEN sample file (", nrow(sample.id)-1, ") does not match number of samples in BGEN file (", bgenInfo$N,")."))
       }

R/glmm.gei.meta.R

@@ -6,7 +6,7 @@ glmm.gei.meta <- function(files, outfile, interaction, SNPID = rep("SNPID", leng
   if(length(Non_Effect_Allele) != k) stop("Error: \"Non_Effect_Allele\" must have the same length as \"files\"!")
   if(length(Effect_Allele) != k) stop("Error: \"Effect_Allele\" must have the same length as \"files\"!")
   col.include <- c("Beta_Marginal", "SE_Beta_Marginal", "P_Value_Marginal", "Beta_G", paste0("Beta_G.",interaction), "SE_Beta_G", paste0("SE_Beta_G.", interaction), paste0("Cov_Beta_G_G.", interaction), "P_Value_Interaction", "P_Value_Joint")
-  master <- read.table(files[1], header=T, as.is=T)[, c(SNPID[1], CHR[1], POS[1],Non_Effect_Allele[1], Effect_Allele[1], "N_Samples", "AF", col.include)]
+  master <- fread(files[1], header=T, data.table = FALSE)[, c(SNPID[1], CHR[1], POS[1],Non_Effect_Allele[1], Effect_Allele[1], "N_Samples", "AF", col.include)]
   names(master)[1:5] <- c("SNPID", "CHR", "POS", "Non_Effect_Allele", "Effect_Allele")
   master <- master[apply(!is.na(master[, col.include]), 1, all),]
   master$SNPID <- paste(master$CHR, master$POS, master$Non_Effect_Allele, master$Effect_Allele, sep = ":")
@@ -23,7 +23,7 @@ glmm.gei.meta <- function(files, outfile, interaction, SNPID = rep("SNPID", leng
   flag <- rep(0, nrow(master))
   if(k > 1) {
     for(i in 2:k) {
-      tmp <- read.table(files[i], header=T, as.is=T)[, c(SNPID[i], CHR[i],POS[i],Non_Effect_Allele[i], Effect_Allele[i], "N_Samples", "AF", col.include)]
+      tmp <- fread(files[i], header=T, data.table = FALSE)[, c(SNPID[i], CHR[i],POS[i],Non_Effect_Allele[i], Effect_Allele[i], "N_Samples", "AF", col.include)]
       names(tmp)[1:5] <- c("SNPID", "CHR", "POS", "Non_Effect_Allele", "Effect_Allele")
       tmp <- tmp[apply(!is.na(tmp[, col.include]), 1, all),]
       tmp$SNPID <- paste(tmp$CHR, tmp$POS, tmp$Non_Effect_Allele, tmp$Effect_Allele, sep = ":")

README.md

@@ -12,7 +12,7 @@ See Section 3.2 of the <a href="https://github.com/large-scale-gxe-methods/MAGEE
 For optimal computational performance, it is recommended to use an R version configured with the Intel Math Kernel Library (or other fast BLAS/LAPACK libraries). See the <a href="https://www.intel.com/content/www/us/en/developer/articles/technical/using-onemkl-with-r.html">instructions</a> on building R with Intel MKL.
 
 ## Version
-The current version is 1.2.0 (June 2, 2022).
+The current version is 1.2.1 (March 23, 2023).
 
 ## License
 This software is licensed under GPL-3.

man/MAGEE.Rd

@@ -12,7 +12,7 @@ Use a glmmkin class object from the null GLMM to perform variant set-based main
 \usage{
 MAGEE(null.obj, interaction, geno.file, group.file,  group.file.sep = "\t", 
       bgen.samplefile = NULL, interaction.covariates = NULL, meta.file.prefix = NULL,
-      MAF.range = c(1e-7, 0.5), MAF.weights.beta = c(1, 25), miss.cutoff = 1, 
+      MAF.range = c(1e-7, 0.5), AF.strata.range = c(0, 1), MAF.weights.beta = c(1, 25), miss.cutoff = 1, 
       missing.method = "impute2mean", method = "davies", tests = "JF", 
       use.minor.allele = FALSE, auto.flip = FALSE, 
       Garbage.Collection = FALSE, is.dosage = FALSE, ncores = 1)
@@ -21,7 +21,7 @@ MAGEE.prep(null.obj, interaction, geno.file, group.file, interaction.covariates
            group.file.sep = "\t", auto.flip = FALSE)
 
 MAGEE.lowmem(MAGEE.prep.obj, geno.file = NULL, meta.file.prefix = NULL, 
-             MAF.range = c(1e-7, 0.5), MAF.weights.beta = c(1, 25), miss.cutoff = 1, 
+             MAF.range = c(1e-7, 0.5), AF.strata.range = c(0, 1), MAF.weights.beta = c(1, 25), miss.cutoff = 1, 
              missing.method = "impute2mean", method = "davies", tests = "JF", 
              use.minor.allele = FALSE, Garbage.Collection = FALSE, is.dosage = FALSE,
 	     ncores = 1)
@@ -54,6 +54,9 @@ the prefix for meta-analysis (default = \code{"NULL"}).
 }
   \item{MAF.range}{
 a numeric vector of length 2 defining the minimum and maximum minor allele frequencies of variants that should be included in the analysis (default = c(1e-7, 0.5)).
+}
+  \item{AF.strata.range}{
+a numeric vector of length 2 defining the minimum and maximum coding allele frequencies of variants in each stratum that should be included in the analysis, if the environmental factor is categorical (default = c(0, 1)).
 }
   \item{MAF.weights.beta}{
 a numeric vector of length 2 defining the beta probability density function parameters on the minor allele frequencies. This internal minor allele frequency weight is multiplied by the external weight given by the group.file. To turn off internal minor allele frequency weight and only use the external weight given by the group.file, use c(1, 1) to assign flat weights (default = c(1, 25)).
@@ -116,10 +119,10 @@ mean coding allele frequency for variants in the test unit group.
 maximum coding allele frequency for variants in the test unit group.
 }
   \item{freq.strata.min}{
-minimum coding allele frequency of each strata if the environmental factor is categorical.
+minimum coding allele frequency of each stratum if the environmental factor is categorical.
 }
   \item{freq.strata.max}{
-maximum coding allele frequency of each strata if the environmental factor is categorical.
+maximum coding allele frequency of each stratum if the environmental factor is categorical.
 }
   \item{MV.pval}{
 MV test p-value.