#/home/moulos/Rbase/bin/R
library(metaseqR)

the.path <- "/data/images/proton/run377/www"
the.contrasts.1 <- c(
	"TAM_vs_T",
	"WU_vs_WTH",	
	"WU_vs_WTR",
	"TH_vs_WTH",
	"TR_vs_WTR"
)

 targets <- read.targets("targets.txt")
 transcript.data <- read.delim("/data/results/tools/rnaseq/metaseqr/mm10/transcript_data_mm10.txt.gz")
rownames(transcript.data) <- as.character(transcript.data$transcript_id)

# metaseqR related variables outside the pipeline
multic <- check.parallel(0.5) # If wish to use multiple cores

 r2c.out <- read2count(targets,transcript.data,has.all.fields=TRUE,multic=multic)

the.counts <- cbind(r2c.out$mergedann,r2c.out$counts)

 metaseqr(
	counts=the.counts,
    sample.list=targets$samples,
    contrast=the.contrasts.1,
    annotation="embedded",
	gene.file="/data/results/tools/rnaseq/metaseqr/mm10/gene_data_mm10.txt.gz",
	id.col=4, # required with embedded annotation
	gc.col=5, # required with embedded annotation
	bt.col=8, # required with embedded annotation
	name.col=7, # required with embedded annotation
    org="custom",
    count.type="utr",
    normalization="deseq", # or whatever supported
    statistics="deseq", # or whatever supported, more than one also
    fig.format=c("png","pdf"),
    export.where=file.path(the.path,"metaseqr_quantseq_run377"),
    restrict.cores=0.5, # fraction of available cores to use
    qc.plots=c(
        "mds","biodetection","countsbio","saturation","readnoise","filtered",
        "correl","pairwise","boxplot","gcbias","lengthbias","meandiff",
        "meanvar","biodist","volcano","deheatmap"
    ),
    exon.filters=NULL,
    gene.filters=list(
       length=list(
            length=500
        ),
        avg.reads=list(
            average.per.bp=100,
            quantile=0.25
        ),
        expression=list(
            median=TRUE,
            mean=FALSE,
            quantile=NA,
            known=NA,
            custom=NA
        ),  # it's the default anyway
        biotype=get.defaults("biotype.filter","mm10")
    ),
    pcut=0.05, # only for the truncated significant list output, all results are exported anyway
    export.what=c("annotation","p.value","adj.p.value","fold.change","stats",
        "counts","flags"), # if you use pandora, the fields "meta.p.value" and "adj.meta.p.value" should be added
    export.scale=c("natural","log2","rpgm"),
    export.values="normalized",
    export.stats="mean"
)