# ploting soliD reads quality
# read quality value distribution for sequencing reads SoliD
# author : Yann Audic
# 2015
# not meant to be run as is
############# SOLiD quality value ##############
# The QV value is calculated using a phred like score
# q = -10 × log10 (p)
# where q is the quality value and (p) is the
# predicted probability that the color call is incorrect 
################################################
# library required
library("bigmemory")

# DIRECTORY CLIPseq
INDIR <- "/private/staff/c/ic/yaudic/Documents/SCIENCE/CLIPCELF1/CLIPSEQ_input"
subdir <- c("AIA0BOLS_reads","AIA0BOLS_20091023_FC1","AIA0BOLS_20091023_FC2","AIA0AOLS_reads","AIA0AOLS_20091023_FC1","AIA0AOLS_20091023_FC2")
# defining the files required
allfiles <- dir(file.path(INDIR,subdir),full.name=TRUE)
# select quality files
qual_files <- allfiles [grep("QV.qual$",allfiles)]
# RNASEQ qual files
RNAseq <- "/private/staff/c/ic/yaudic/Documents/SCIENCE/CLIPCELF1/RNAseq_input/NICKEL_20100706_FC2/NICKEL_20100706_FC2_AIA_COLS1_F3_QV.qual"
qual_files <- c(qual_files,RNAseq)

## function definition

##convert_comment
# function to replace the > of the quality file by the comment character
# this will allow the subsequent input of the qual file
# the return value is the full path of the modified file
# the transitory file outfile1 is removed

convert_comment<-function(infile)
{
	x<-infile
	outfile1<-paste(infile,".out.qual",sep="")
# using sed to replace > by # sed 's/>/#/' *.qual > *.out.qual
	exe <- paste("sed 's/>/#/' ",infile,sep="")
	exe <- paste(exe, outfile1, sep= " > ")
# convert > to #
	system(exe)

	outfile2 <- paste(infile,".trimmed.qual",sep="")
# using sed to delete comment lines # sed 's/>/#/' *.out.qual > *.trimmed.qual
	exe <- paste("sed '/#/d' ",outfile1,sep="")
	exe <- paste(exe, outfile2, sep= " > ")
# delete all comment lines
	system(exe)
	file.remove(outfile1)
return <- outfile2
}

# usage value=convert_comment(infile)
## end convert_comment

## big.bxplot.stats
# function to compute boxplot stats from a bigfile

# we can calculate the boxplot stats for each column with
# boxplot.stats(x, coef = 1.5, do.conf = TRUE, do.out = TRUE)
# to plot a boxplot we need the stats and the n

big.bxplot.stats<-function(data_bxp)
{
# number of colum for boxplot
	num_col <- dim(data_bxp)[2]
# variable definition
	bxp.stat <- list()
	stats.mat <- matrix(nrow=5,ncol=num_col)
	number <- c()

for (i in c(1:num_col))
	{
	bxp.stat[[i]] <- boxplot.stats(w[,i], coef = 1.5, do.conf = TRUE, do.out = TRUE)
	stats.mat[,i] <- bxp.stat[[i]]$stats
	number <- c(number,bxp.stat[[i]]$n)
	}
# we return only the values required to generate boxplot with bxp
return <- list(stats=stats.mat, n=number)
}

## end big.bxplot.stats

############ ANALYSIS ###########
# file backing is required for such a big file

w=read.big.matrix(paste(qual_files[2],".trimmed.qual",sep=""),col.names = paste("N",c(1:50),sep=" "),sep=" ",type ="integer",backingfile="W",backingpath="/private/staff/c/ic/yaudic/temp")

statistics <- list()
length(statistics) <- 7 # 7 files to process
for (i in c(1:7))
{
rm(w)
name<-convert_comment(qual_files[i]) # take some time for big files (7.5 Go)
# name=paste(qual_files[1],".trimmed.qual",sep="")
w <-read.big.matrix(name,col.names = paste("N",c(1:50),sep=" "),sep=" ",type ="integer",backingfile=paste("W",i,sep="_"),backingpath="/private/staff/c/ic/yaudic/temp") # take some time also
statistics[[i]]<-big.bxplot.stats(w)
}

# create boxplot
for (i in c(1:7))
{
svg(filename=file.path(OUTDIR,paste(i,"quality.svg",sep="_")))
bxp(statistics[[i]])
dev.off()
}

######### END ANALYSIS #########

######### ENVIRONMENT ##########
#R version 3.2.0 (2015-04-16)
#Platform: x86_64-suse-linux-gnu (64-bit)
#Running under: openSUSE 13.1 (Bottle) (x86_64)#

#locale:
# [1] LC_CTYPE=fr_FR.UTF-8       LC_NUMERIC=C              
# [3] LC_TIME=fr_FR.UTF-8        LC_COLLATE=fr_FR.UTF-8    
# [5] LC_MONETARY=fr_FR.UTF-8    LC_MESSAGES=fr_FR.UTF-8   
# [7] LC_PAPER=fr_FR.UTF-8       LC_NAME=C                 
# [9] LC_ADDRESS=C               LC_TELEPHONE=C            
#[11] LC_MEASUREMENT=fr_FR.UTF-8 LC_IDENTIFICATION=C       
#
#attached base packages:
#[1] stats     graphics  grDevices utils     datasets  methods   base     

#other attached packages:
#[1] bigmemory_4.5.8     bigmemory.sri_0.1.3

#loaded via a namespace (and not attached):
#[1] Rcpp_0.12.1