improved master template

README.md

@@ -25,6 +25,7 @@ To use the structure from above when working on bioinformatics-srv1 just use
 /home/brandl/mnt/mack/bioinfo/scripts/ngs_tools/v1.0
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 
+
 ToDo
 ====
 

chipseq_workflow/create_igv_session.R

@@ -3,9 +3,11 @@
 #' # Create an IGV Session File
 #' ## Create an IGV Session File
 #' hallo
-
+#http://www.slideshare.net/EdwindeJonge1/docopt-user2014
 suppressMessages(library(docopt))
 
+#doc <- 'Usage: igvtrack.R --genome [options] <genome> <bam_files>...
+
 doc <- 'Usage: igvtrack.R --genome [options] <genome> <bam_files>...
 
 Options:

dge_workflow/dge_master_template.sh

-export baseDir=<<PATH_TO_BASEDIR>>
+# todo define project name
 export project=<<PROJECTNAME>>
 # screen -R $project
 
 
-## intially
-DGE_HOME=/projects/bioinfo/holger/bioinfo_templates/dge_workflow
-## when customization is needed
-#export DGE_HOME=/sw/users/brandl/projects/<<PROJECTNAME>>
+## madmax
+if [ -n "$LSF_SERVERDIR" ]; then
+    export baseDir="/projects/bioinfo/holger/projects/${project}"
+    export PROJECT_SCRIPTS="/projects/bioinfo/holger/scripts/${project}"
+    export DGE_HOME="/projects/bioinfo/scripts/ngs_tools/v1.0/"
+fi
 
-source $DGE_HOME/dge_utils.sh
-export PATH=$DGE_HOME:$DGE_HOME/../misc/:$PATH
+## bioinfo
+if [ $(hostname) == "bioinformatics-srv1" ]; then
+    #<<<todo define paths on bioinfo>>>
+    #export baseDir=/home/brandl/mnt/chip-seq_study/ChIPSeq_March_2015/data
+    #export PROJECT_SCRIPTS=/home/brandl/mnt/chip-seq_study/ChIPSeq_March_2015/scripts
+    export DGE_HOME="/home/brandl/mnt/mack/bioinfo/scripts/ngs_tools/v1.0"
+fi
 
 
+source $DGE_HOME/dge_workflow/dge_utils.sh
+export PATH=$DGE_HOME/dge_workflow:$PATH
+
+
+## todo define igenome to be used
+## igenome=/projects/bioinfo/igenomes/Canis_familiaris/Ensembl/CanFam3.1
+igenome=<<<<TBD>>>>
+
+########################################################################################################################
+### Fetch the data
+
+mcdir $baseDir/originals
+
+wget -nc --user="USER" --password="PW"  -r --no-directories  --no-check-certificate -A "*fastq.gz" https:/projects.biotec.tu-dresden.de/ngs-filesharing/martaf/
+
+mailme "$project: fastq download done"
+
+
+## todo make sure to also copy the sample sheet in here
+
 ########################################################################################################################
 ### QC
 
 dge_fastqc $(ls *fastq.gz) &
 
+########################################################################################################################
+### Apply renaming and merge lane replicates (but keep technical ones)
+
+## todo adjust renaming scheme to project specifics
+
+mcdir $baseDir/lanereps_pooled
+
+echo '
+devtools::source_url("https://dl.dropboxusercontent.com/u/113630701/datautils/R/core_commons.R")
+options(java.parameters = "-Xmx4g" ); library(xlsx)
+
+#sheetFile <- "/projects/bioinfo/holger/projects/stefania_isnm1/originals/stefaniat-FC_SN678_293-2014-11-26.xls"
+sheetFile <- commandArgs(T)[1]
+
+sampleSheet <- read.xlsx2(sheetFile, "Fastqfiles") %>%
+    select(File, SampleID, SampleName) %>%
+    ## clean up the formatting a bit
+    mutate(
+        SampleName=str_replace(SampleName, "Ctrl14", "Ctrl_14"),
+        SampleName=ifelse(str_detect(SampleName, "_R"), SampleName, paste0(SampleName, "_R1")),
+        SampleName=str_replace(SampleName, "_14", "_"),
+        SampleName=str_replace(SampleName, "_R", "_TR")
+    ) %>%
+    ## also a add a column containing just biological replicate info (for bam merging later)
+    mutate(
+        bio_sample=str_replace(SampleName, "_TR(1|2)", "")
+    )
+
+write.delim(sampleSheet, file="renaming_scheme.txt")
+
+require(ggplot2)
+#ggplot(sampleSheet, aes(SampleName)) + geom_bar() + coord_flip()
+## merge lane replication
+
+library(foreach)
+library(doMC)
+registerDoMC(cores=4)
+
+#sampleSheet %>% group_by(SampleName) %>% do(with(., print(paste("zcat", paste(File, collapse=" "), "| gzip -c >", paste0(SampleName, ".fastq.gz")))))
+## with parallelization
+sampleSheet %>% d_ply(.(SampleName), with,
+    system(paste("zcat", paste(paste0("../originals/", File), collapse=" "), "| gzip -c >", paste0(SampleName, ".fastq.gz")))
+, .parallel=T)
+' | R --vanilla -q --args $baseDir/originals/stefaniat-FC_SN678_293-2014-11-26.xls
+
+
 
 ########################################################################################################################
 ### Trim low-quality bases and remove left-over adapters
 
 mcdir $baseDir/trimmed
 
-dge_cutadapt $(ls $baseDir/treps_pooled/*fastq.gz) 2>&1 | tee cutadapt.log
+dge_cutadapt $(ls $baseDir/lanereps_pooled/*fastq.gz) 2>&1 | tee cutadapt.log
 
 dge_fastqc $(ls *fastq.gz) &
 
@@ -34,10 +107,6 @@ dge_fastqc $(ls *fastq.gz) &
 mcdir $baseDir/mapped
 
 fastqFiles=$(ls $baseDir/trimmed/*.fastq.gz)
-igenome=<<<<TBD>>>>
-## Example:
-## igenome=/projects/bioinfo/igenomes/Canis_familiaris/Ensembl/CanFam3.1
-
 
 dge_tophat_se -i $igenome $fastqFiles 2>&1 | tee mapped.log
 
@@ -67,25 +136,32 @@ gtfFile=$igenome/Annotation/Genes/genes.gtf
 
 ## define labels to split bam files into replicate groups
 #labels=$(csvcut -t -c1 $baseDir/lanereps_pooled/sample2replicates.txt | tail -n+2 | uniq | xargs echo | sed 's/ /,/g')
+## better externalize them
 labels=<<<<TBD>>>>
 
 dge_cuffdiff $gtfFile $baseDir/mapped $labels
 
 mailme "$project: cuffdiff done"
 
-
 mcdir $baseDir/cuffdiff/dge_report
-#export DGE_PARAMS="-S"
-spin.R $DGE_HOME/dge_analysis.R $baseDir/cuffdiff
+
+spin.R $DGE_HOME/cuffdiff_report.R ..
+
+## or when using specfic contrasts
+#echo "Ctrl_12h,X11B_12h
+#Ctrl_36h,X11B_36h" > contrasts.txt
+#spin.R $DGE_HOME/cuffdiff_report.R  \""--constrasts contrasts.txt --pcutoff 0.01 $baseDir/cuffdiff"\"
+
 
 
 ########################################################################################################################
 ### Sync back to project space
 
-# ... project specific stuff
+# uni
 screen -R rsync_$project
 
 ## main sync
+## todo define mount path on bioinfo for bidirectional synching
 ~/bin/unison $baseDir ssh://bioinfo///home/brandl/mnt/<<MOUNT_PATH>> -fastcheck true -times -perms 0