first commit

codescript/EPIC.modified.R

@@ -0,0 +1,94 @@
+suppressPackageStartupMessages(library("optparse"))
+suppressPackageStartupMessages(library("stats"))
+
+# specify our desired options in a list
+# by default OptionParser will add an help option equivalent to 
+# make_option(c("-h", "--help"), action="store_true", default=FALSE, 
+#               help="Show this help message and exit")
+
+option_list <- list( 
+  make_option(c("-p", "--prefix"), type="character",default="./",
+              help="The output files prefix [default ./]"),
+  make_option(c("-i", "--input"),type="character", default=NULL,
+              help="The directory of input EPIC files. required!")
+)
+
+# get command line options, if help option encountered print help and exit,
+# otherwise if options not found on command line then set defaults, 
+opt <- parse_args(OptionParser(option_list=option_list))
+
+if (is.null(opt$input)){
+  print_help(opt_parser)
+  stop("At least one argument must be supplied (input file).", call.=FALSE)
+}
+
+# load libraries
+
+library("minfi")
+library("IlluminaHumanMethylationEPICmanifest")
+library("IlluminaHumanMethylationEPICanno.ilm10b4.hg19")
+
+## 1. data import
+targets <- read.metharray.sheet(opt$input)
+rgSet <- read.metharray.exp(targets = targets)
+targets$ID <- paste(targets$Sample_Group,targets$Sample_Name,sep=".")
+sampleNames(rgSet) <- targets$ID
+#phenoData <- pData(rgSet)
+message("data import:finished")
+
+## 2. Quality control
+# 2.1 qc report by minfi
+# qcReport(rgSet, sampNames=targets$ID, sampGroups=targets$Sample_Group,pdf="qcReport.pdf")
+          
+# 2.2 data filtering
+# get detected p value
+# a. remove samples with average p value less than 0.05
+detP <- detectionP(rgSet)
+keep <- colMeans(detP) < 0.05
+rgSet_sample_removed <- rgSet[,keep]
+targets_sample_removed <- targets[keep,]
+message("sample p value filtration:finished")
+
+# b. normalization
+mSetSq <- preprocessFunnorm(rgSet_sample_removed)
+message("Funnorm normalization:finished")
+
+# c. filter probes with p value less than 0.01
+# ensure probes are in the same order in the mSetSq and detP objects
+detP <- detP[match(featureNames(mSetSq),rownames(detP)),] 
+# remove any probes that have failed in one or more samples
+keep <- rowSums(detP < 0.01) > ncol(mSetSq) * 0.9 
+mSetSqFlt <- mSetSq[keep,]
+message("probe p value filteration:finished")
+
+# d. remove sex probes
+annotation <- getAnnotation(IlluminaHumanMethylationEPICanno.ilm10b4.hg19)
+keep <- !(featureNames(mSetSqFlt) %in% annotation$Name[annotation$chr %in% c("chrX","chrY")])
+mSetSqFlt <- mSetSqFlt[keep,]
+mSetSqFlt <- dropLociWithSnps(mSetSqFlt)
+message("remove:finished")
+
+## get M and beta value
+# filtered
+mVals <- getM(mSetSqFlt)
+bVals <- getBeta(mSetSqFlt)
+# raw
+mSetRaw <- preprocessRaw(rgSet)
+raw_mVals <- getM(mSetRaw)
+raw_bVals <- getBeta(mSetRaw)
+message("m value and beta value output:finished")
+
+# write output
+message("saving R data")
+rdata_filename = paste(opt$prefix, '.RData',sep="")
+save(rgSet, targets, detP, file = rdata_filename)
+message("writing filtered table")
+m_filename = paste(opt$prefix,'.filter.p.sex.snp.mVal.txt',sep="")
+write.table(mVals,m_filename,col.names = T,row.names = T,sep="\t",quote=F)
+b_filename = paste(opt$prefix,'.filter.p.sex.snp.bVal.txt',sep="")
+write.table(bVals,b_filename,col.names = T,row.names = T,sep="\t",quote=F)
+message("writing raw table")
+m_raw_filename = paste(opt$prefix,'.raw.mVal.txt',sep="")
+write.table(raw_mVals,m_raw_filename,col.names = T,row.names = T,sep="\t",quote=F)
+b_raw_filename = paste(opt$prefix,'.raw.bVal.txt',sep="")
+write.table(raw_bVals,b_raw_filename,col.names = T,row.names = T,sep="\t",quote=F)

inputs

@@ -0,0 +1,7 @@
+{
+  "{{ project_name }}.disk_size": "200",
+  "{{ project_name }}.prefix": "{{ prefix }}",
+  "{{ project_name }}.docker": "registry-vpc.cn-shanghai.aliyuncs.com/pgx-docker-registry/r-base:v1.0",
+  "{{ project_name }}.cluster_config": "OnDemand bcs.b4.xlarge img-ubuntu-vpc",
+  "{{ project_name }}.input_dir": "{{ input_dir }}"
+}

readme.md

@@ -0,0 +1,32 @@
+## minfi分析Illumina 850K(EPIC)
+
+#APP介绍
+
+###甲基化原理简述
+
+Illumina 850K甲基化芯片可同时检测>850,000个位点，覆盖>95%的CpG岛，99%的RefSeq基因，已经成为精准医学研究的重要方法之一。
+
+850K芯片采用了两种探针Infinium Ⅰ 和Infinium Ⅱ对样品甲基化进行测定，Infinium I采用了两种bead（甲基化M和非甲基化U，如图显示），而II只有一种bead（即甲基化和非甲基化在一起），这也导致了它们在后续荧光探测的不同，而根据不同探针的bead的荧光值，就可以得到样品各个位点上的甲基化水平。
+
+###APP简介
+
+为了更好更便捷的分析全基因组甲基化数据，我们选用了分析850K芯片的R包——minfi包，构建了分析pipeline，可以得到全基因组的各个位点甲基化表达谱。
+
+#流程和参数
+
+#### 850K array分析流程
+
+#输入和输出
+
+###输入
+
+需要一个文件夹，其中包含：
+
+▪   idat文件：样本的芯片测序结果文件，命名方式为：“'Sentrix_ID'_'Sentrix_Position'__Grn/Red.idat”
+
+▪   sample_sheet文件：样本的注释信息文件，命名为“sample_sheet.csv”，其中包含Sample_Name，Sentrix_ID, Sentrix_Position, Sample_Group等注释信息
+
+###输出
+
+在850K芯片的分析中，beta 值是最常用的甲基化水平的定量方式，其主要用于差异分析。
+

tasks/minfi.wdl

@@ -0,0 +1,28 @@
+task minfi {
+    File prefix
+    File input_dir
+    String docker
+    String cluster_config
+    String disk_size
+
+    command <<<
+		  set -o pipefail
+		  set -e	
+      Rscript /opt/EPIC.modified.R -p ${prefix} -i ${input_dir}
+    >>>
+
+    runtime {
+		  docker:docker
+		  cluster: cluster_config
+      systemDisk: "cloud_ssd 40"
+      dataDisk: "cloud_ssd " + disk_size + " /cromwell_root/"
+    }
+
+    output {
+		  File rdata = "${prefix}.RData"
+		  File filter_mVal = "${prefix}.filter.p.sex.snp.mVal.txt"
+		  File filter_bVal = "${prefix}.filter.p.sex.snp.bVal.txt"
+      File raw_mVal = "${prefix}.raw.mVal.txt"
+      File raw_bVal = "${prefix}.raw.bVal.txt"
+    }
+  }

workflow.wdl

@@ -0,0 +1,19 @@
+import "./tasks/minfi.wdl" as minfi
+
+workflow {{ project_name }} {
+  File prefix
+  File input_dir
+  String docker
+  String cluster_config
+  String disk_size
+
+	call minfi.minfi as minfi {
+		input:
+		prefix=prefix,
+		input_dir=input_dir,
+		docker=docker,
+		cluster_config=cluster_config,
+		disk_size=disk_size
+	}
+}
+