# TODO  define project name
export project="TODO  define project name"
# screen -R ${project}


## madmax
if [ "$HOSTNAME"=="falcon1" ]; then
    export baseDir="/projects/bioinfo/holger/projects/${project}"
    export PROJECT_SCRIPTS="/projects/bioinfo/holger/scripts/${project}"
    export NGS_TOOLS="/projects/bioinfo/scripts/ngs_tools/dev"
fi

## bioinfo
if [ $(hostname) == "bioinformatics-srv1" ]; then
    export baseDir=/net/mack/lustre/projects/bioinfo/holger/projects/${project}
    export PROJECT_SCRIPTS==/net/mack/lustre/projects/bioinfo/holger/scripts/${project}
    export NGS_TOOLS=/net/mack/lustre/projects/bioinfo/scripts/ngs_tools/dev
fi


source ${NGS_TOOLS}/dge_workflow/dge_utils.sh
export PATH=${NGS_TOOLS}/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"

### Basic QC
dge_fastqc $(ls *fastq.gz) &

## todo make sure to also copy the sample sheet in here

########################################################################################################################
### 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://raw.githubusercontent.com/holgerbrandl/datautils/v1.36/R/core_commons.R")

sheetFile <- "../originals/natalied-FC_SN678_338-2015-5-12.xls"

sampleSheet <- read_excel(sheetFile, "Fastqfiles") %>%
    select(File, SampleName) %>%
    mutate(
        bio_replicate=str_match(SampleName, "(.).*")[,2],
        sample = str_replace(SampleName, "[0-9]*", "") %>% str_replace_all(c("NC" = "no_culture", "NA" = "no_hormone", "ECD"="ecdysone_", "INS"="insulin_", "9"="9h", "4"="4h")),
        bio_sample=paste(sample, bio_replicate, sep="_")
    )

write_tsv(sampleSheet, path="renaming_scheme.txt")

require(ggplot2)
ggplot(sampleSheet, aes(bio_sample)) + geom_bar() + coord_flip()

sampleSheet %>% group_by(bio_sample) %>% summarise(
        zcat=paste("zcat", paste(paste0("../originals/", File), collapse=" "), "| gzip -c >", paste0(bio_sample[1], ".fastq.gz"))
    ) %$%
    zcat %>%
    write_lines("lane_merge.cmd")

sampleSheet %>% distinct(replicate, sample) %>%
    arrange(sample, replicate) %>%
    write_tsv(path="basic_design.txt")
' | R --vanilla -q

cat lane_merge.cmd | while read line; do
#    eval ${line}
    jl submit -j .repmerge "$line"
done

jl wait --email --report

dge_fastqc $(ls *fastq.gz) &


########################################################################################################################
### Alignment the reads

mcdir $baseDir/alignments

star_align.kts ${igenome} $(ls ${baseDir}/lanereps_pooled/*.fastq.gz) 2>&1 | tee star_algin.log

#dge_bam_correlate . & # part of star_align.kts now

mailme "$project: mapping done"


########################################################################################################################
### Differential Expression Analysis


mcdir $baseDir/dge_analysis

## build custom design matrix
#csvcut -tc replicate,sample $baseDir/lanereps_pooled/renaming_scheme.txt |  csvformat -T > design_matrix.txt
#Rscript -  <<"EOF"
#devtools::source_url("https://raw.githubusercontent.com/holgerbrandl/datautils/v1.36/R/core_commons.R")
#file.path(Sys.getenv("baseDir"), "lanereps_pooled/renaming_scheme.txt") %>%
#    read_tsv() %>%
#    distinct(replicate, sample) %>%
#    arrange(sample, replicate) %>%
#    write_tsv("design_matrix.txt")
#EOF



## or use just some contrasts of interest
#echo "
#sample_1, sample_2
#unpolarised,liver_polar_stage3
#" | trim | csvformat -T  > contrasts.txt
#


rend.R -e ${NGS_TOOLS}/dge_workflow/featcounts_deseq_mf.R ../alignments/star_counts_matrix.txt  2>&1  | tee featcounts_deseq.log

#rend.R -e ${NGS_TOOLS}/dge_workflow/featcounts_deseq_mf.R --design 'batch+sample' --pcutoff 0.01 --contrasts ../contrasts_wt.txt ../qseq.counts.txt ../qseq.design_matrix.txt 2>&1  | tee featcounts_deseq_mf.log



## Term enrichment
mcdir ${baseDir}/dge_analysis/dge_enrichment_analysis
#$NGS_TOOLS/common/cp_enrichment.R --overlay_expr_data ../plot_score_matrix.txt ../degs_by_contrast.txt contrast
rend.R -e ${NGS_TOOLS}/common/cp_enrichment.R --overlay_expr_data ../plot_score_matrix.txt ../degs_by_contrast.txt contrast


########################################################################################################################
### Sync back to project space

## version common tools
## todo make sure to stay in current directory #{ (cd ${NGS_TOOLS} && git describe --tags) } >> .used_ngs_tools
(cd ${NGS_TOOLS} && test -n "$project"   && git diff --exit-code && git tag "${project}__$(date +'%Y%m%d')")  || echo "could not tag current branch"
# and log it
git describe --tags >> ${baseDir}/.used_ngs_tools



## bidirectional sync with project space
#remoteProjectDir=/net/fileserver-nfs/stornext/snfs3/projects/ TODO define mount path on bioinfo for bidirectional synching
(ls "${remoteProjectDir}" "${baseDir}" && ~/bin/unison ${baseDir}  ${remoteProjectDir}/data -fastcheck -times -perms 0 -batch) || echo "could not sync data"

## update scripts working copy in project space
#git clone "git@git.mpi-cbg.de:bioinfo/${project}.git" ${remoteProjectDir}/scripts
(ls "${remoteProjectDir}/scripts" > /dev/null && cd ${remoteProjectDir}/scripts/ && git pull) || echo "could not sync data"


## also update the git copy of the script


# or use a uni-directional sync
#rsync -avsn --delete  ${baseDir} brandl@fileserver:/projects//file/server/path
mailme "$project: sync done"