Merge pull request #558 from belleau/main

Add function processBlockChr and correct some importFrom

NAMESPACE

@@ -2,6 +2,7 @@
 
 export(add1KG2SampleGDS)
 export(addGeneBlockGDSRefAnnot)
+export(addGeneBlockRefAnnot)
 export(addRef2GDS1KG)
 export(addStudy1Kg)
 export(computeAncestryFromSynthetic)

R/process1KG.R

@@ -544,7 +544,8 @@ generatePhase1KG2GDS <- function(gdsReference, gdsReferencePhase,
 #' unlink(fileRefPhaseGDS, force=TRUE)
 #'
 #' @author Pascal Belleau, Astrid Deschênes and Alexander Krasnitz
-#' @importFrom gdsfmt index.gdsn read.gdsn readmode.gdsn
+#' @importFrom gdsfmt createfn.gds closefn.gds index.gdsn read.gdsn readmode.gdsn
+#' @importFrom SNPRelate snpgdsOpen
 #' @encoding  UTF-8
 #' @export
 generatePhaseRef <- function(fileReferenceGDS, fileReferenceAnnotGDS,
@@ -774,6 +775,8 @@ identifyRelative <- function(gds, maf=0.05, thresh=2^(-11/2),
 #' @author Pascal Belleau, Astrid Deschênes and Alexander Krasnitz
 #'
 #' @importFrom GENESIS pcairPartition
+#' @importFrom gdsfmt closefn.gds
+#' @importFrom SNPRelate snpgdsOpen
 #' @importFrom S4Vectors isSingleNumber
 #' @importFrom methods is
 #' @encoding UTF-8
@@ -997,7 +1000,7 @@ getRef1KGPop <- function(gdsReference, popName="superPop") {
 #'
 #'
 #' @author Pascal Belleau, Astrid Deschênes and Alexander Krasnitz
-#' @importFrom gdsfmt index.gdsn read.gdsn
+#' @importFrom gdsfmt index.gdsn read.gdsn closefn.gds
 #' @importFrom stats rmultinom
 #' @importFrom SNPRelate snpgdsOpen
 #' @encoding UTF-8
@@ -1140,3 +1143,95 @@ addGeneBlockGDSRefAnnot <- function(gdsReference, gdsRefAnnotFile,
     return(0L)
 }
 
+#' @title Append information associated to blocks, as indexes, into the
+#' Population Reference SNV Annotation GDS file
+#'
+#' @description The function appends the information about the blocks into
+#' the Population Reference SNV Annotation GDS file. The information is
+#' extracted from the Population Reference GDS file.
+#'
+#' @param fileReferenceGDS  a \code{character} string representing the file
+#' name of the Reference GDS file. The file must exist.
+#'
+#' @param gdsRefAnnotFile a \code{character} string representing the
+#' file name corresponding the Reference SNV
+#' Annotation GDS file. The function will
+#' open it in write mode and close it after. The file must exist.
+#'
+#' @param winSize a single positive \code{integer} representing the
+#' size of the window to use to group the SNVs when the SNVs are in a
+#' non-coding region. Default: \code{10000L}.
+#'
+#' @param ensDb An object with the ensembl genome annotation
+#' Default: \code{EnsDb.Hsapiens.v86}.
+#'
+#' @param suffixBlockName a \code{character} string that identify the source
+#' of the block and that will be added to the block description into
+#' the Reference SNV Annotation GDS file, as example: Ensembl.Hsapiens.v86.
+#'
+#' @return The integer \code{OL} when the function is successful.
+#'
+#' @examples
+#'
+#' ## Path to the demo pedigree file is located in this package
+#' dataDir <- system.file("extdata", package="RAIDS")
+#'
+#  ## Temporary file
+#' fileAnnotGDS <- file.path(tempdir(), "ex1_good_small_1KG_Ann_GDS.gds")
+#'
+#' ## Required library
+#' if (requireNamespace("EnsDb.Hsapiens.v86", quietly=TRUE)) {
+#'
+#'     file.copy(file.path(dataDir, "tests",
+#'         "ex1_NoBlockGene.1KG_Annot_GDS.gds"), fileAnnotGDS)
+#'
+#'     ## Making a "short cut" on the ensDb object
+#'     edb <- EnsDb.Hsapiens.v86::EnsDb.Hsapiens.v86
+#'
+#'     ## GDS Reference file
+#'     fileReferenceGDS  <- file.path(dataDir, "tests",
+#'                 "ex1_good_small_1KG.gds")
+#'
+#'     \donttest{
+#'
+#'
+#'         ## Append information associated to blocks
+#'         addGeneBlockRefAnnot(fileReferenceGDS=fileReferenceGDS,
+#'             gdsRefAnnotFile=fileAnnotGDS,
+#'             ensDb=edb,
+#'             suffixBlockName="EnsDb.Hsapiens.v86")
+#'
+#'         gdsAnnot1KG <- openfn.gds(fileAnnotGDS)
+#'         print(gdsAnnot1KG)
+#'         print(read.gdsn(index.gdsn(gdsAnnot1KG, "block.annot")))
+#'
+#'        closefn.gds(gdsAnnot1KG)
+#'     }
+#'
+#'     ## Remove temporary file
+#'     unlink(fileAnnotGDS, force=TRUE)
+#'
+#' }
+#'
+#' @author Pascal Belleau, Astrid Deschênes and Alexander Krasnitz
+#' @importFrom gdsfmt openfn.gds closefn.gds
+#' @importFrom SNPRelate snpgdsOpen
+#' @importFrom S4Vectors isSingleNumber
+#' @encoding UTF-8
+#' @export
+addGeneBlockRefAnnot <- function(fileReferenceGDS, gdsRefAnnotFile,
+                                    winSize=10000, ensDb, suffixBlockName) {
+
+
+    if (!(is.character(fileReferenceGDS) && (file.exists(fileReferenceGDS)))) {
+        stop("The \'fileReferenceGDS\' must be a character string ",
+             "representing the Reference GDS file. The file must exist.")
+    }
+
+    gdsReference <- snpgdsOpen(filename=fileReferenceGDS)
+    res <- addGeneBlockGDSRefAnnot(gdsReference, gdsRefAnnotFile,
+                                   winSize=10000, ensDb, suffixBlockName)
+    closefn.gds(gdsReference)
+    ## Success
+    return(res)
+}

R/tools_internal.R

@@ -438,3 +438,111 @@ readSNVVCF <- function(fileName,
 
     return(matSample)
 }
+
+#' @title The function create a vector of integer representing the linkage
+#' disequilibrium block for each SNV in the in the same order
+#' than the variant in Population reference dataset.
+#'
+#' @description The function create a vector of integer representing the linkage
+#' disequilibrium block for each SNV in the in the same order
+#' than the variant in Population reference dataset.
+#'
+#' @param fileReferenceGDS  a \code{character} string representing the file
+#' name of the Reference GDS file. The file must exist.
+#'
+#' @param fileBlock a \code{character} string representing the file
+#' name of output file det from the plink block command for a chromosome.
+#'
+#' @return a \code{list} containing 2 entries:
+#' \describe{
+#' \item{\code{chr}}{ a \code{integer} representing a the chromosome from
+#' fileBlock.
+#' }
+#' \item{\code{block.snp}}{ the a \code{array} of integer
+#' representing the linkage disequilibrium block for
+#' each SNV in the in the same order than the variant
+#' in Population reference dataset.
+#' }
+#' }
+#'
+#'
+#' @examples
+#'
+#' ## Path to the demo pedigree file is located in this package
+#' dataDir <- system.file("extdata", package="RAIDS")
+#'
+#' ## Demo of Reference GDS file containing reference information
+#' fileReferenceGDS <- file.path(dataDir, "PopulationReferenceDemo.gds")
+#'
+#' ## Demo of of output file det from the plink block
+#' ## command for chromosome 1
+#' fileLdBlock <- file.path(dataDir, "block.sp.EUR.Ex.chr1.blocks.det")
+#'
+#' listLdBlock <- RAIDS:::processBlockChr(fileReferenceGDS, fileLdBlock)
+#'
+#' @author Pascal Belleau, Astrid Deschênes and Alexander Krasnitz
+#' @importFrom gdsfmt read.gdsn index.gdsn closefn.gds
+#' @importFrom SNPRelate snpgdsOpen
+#' @encoding UTF-8
+#' @keywords internal
+processBlockChr <- function(fileReferenceGDS, fileBlock) {
+
+    if (!(is.character(fileReferenceGDS) && (file.exists(fileReferenceGDS)))) {
+        stop("The \'fileReferenceGDS\' must be a character string ",
+             "representing the Reference GDS file. The file must exist.")
+    }
+    if (!(is.character(fileBlock) && (file.exists(fileBlock)))) {
+        stop("The \'fileBlock\' must be a character string ",
+             "representing the file .det from plink block result. The file must exist.")
+    }
+
+    gdsReference <- snpgdsOpen(filename=fileReferenceGDS)
+    blockChr <- read.delim(fileBlock, sep="")
+
+    listChr <- unique(blockChr$CHR)
+    if(length(listChr) != 1){
+        stop(paste0("There is not just one CHR in ", fileBlock, "\n"))
+    }
+    listChr <- as.integer(gsub("chr", "", listChr))
+    listSNVChr <- read.gdsn(index.gdsn(gdsReference, "snp.chromosome"))
+    listSNVChr <- which(listSNVChr == listChr)
+    snp.keep <- read.gdsn(index.gdsn(gdsReference, "snp.position"))[listSNVChr]
+    closefn.gds(gdsReference)
+    z <- cbind(c(blockChr$BP1, snp.keep, blockChr$BP2+1),
+               c(seq_len(nrow(blockChr)),
+                 rep(0, length(snp.keep)), -1*seq_len(nrow(blockChr))))
+
+    z <- z[order(z[,1]),]
+    block.snp <- cumsum(z[,2])[z[,2] == 0]
+
+    curStart <- 0
+    activeBlock <- 0
+    blockState <- 0
+    block.inter <- rep(0, length(which(block.snp == 0)))
+    k <- 1
+    for(i in seq_len(length(block.snp))){
+        if(block.snp[i] == 0){
+            if(activeBlock == 1){
+                if(snp.keep[i] - curStart >= 10000) {
+                    blockState <- blockState - 1
+
+                    curStart <- snp.keep[i]
+                }
+            } else{
+                blockState <- blockState - 1
+                curStart <- snp.keep[i]
+                curStart <- snp.keep[i]
+                activeBlock <- 1
+            }
+            block.inter[k] <- blockState
+            k <- k + 1
+        }else{
+            activeBlock <- 0
+        }
+    }
+    block.snp[block.snp == 0] <- block.inter
+    res <- list(chr=listChr,
+                block.snp=block.snp)
+    return(res)
+}
+

inst/extdata/block.sp.EUR.Ex.chr1.blocks.det

@@ -0,0 +1,10 @@
+ CHR          BP1          BP2           KB  NSNPS SNPS
+   1        51897        51927        0.031      2 s3|s4
+   1        54707        54715        0.009      2 s6|s7
+   1        55544        59039        3.496      2 s14|s15
+   1        61986        66506        4.521      3 s17|s18|s31
+   1        76837        77873        1.037      3 s39|s42|s43
+   1        79771        80140         0.37      2 s48|s49
+   1        82675        86330        3.656     12 s55|s58|s61|s62|s66|s67|s68|s71|s72|s75|s78|s79
+   1        87189        88337        1.149      3 s81|s89|s96
+   1       631489       633328         1.84      3 s150|s159|s160