finish_process_flowcell

set -euo pipefail

export FLOWCELL=$(echo $(pwd) | cut -d "/" -f5);echo $FLOWCELL
export FLOWCELL_PATH=$(pwd);echo $FLOWCELL_PATH

#move all clean bams from each sample to one directory
mkdir -p all_samples
for i in Sample*;do mv "$i"/uBAM/"$i"_clean.bam ${FLOWCELL_PATH}/all_samples/ ;done ;wait
cd ${FLOWCELL_PATH}/all_samples/
ALL_SAMPLEPATH=$(pwd)

#create a dynamic array to specify all bams as input for Markdups
for i in *_clean.bam;do echo "INPUT="${ALL_SAMPLEPATH}/"$i" >> all_samples.txt ;done ;wait
readarray -t bam_array < all_samples.txt
rm all_samples.txt
export ${bam_array[@]}
echo ${bam_array[@]}


#run markduplicates on all samples. This will remove all duplicates while also merging all the bams into a single file
echo "Step6: Processing MarkDuplicates ALL SAMPLES!"
export FLOWCELL=$(echo $(pwd) | cut -d "/" -f5);echo $FLOWCELL
(java -verbose:gc -XX:+UseParallelOldGC -XX:+AggressiveOpts -XX:ParallelGCThreads=56 -Xmn30g  -Xms60g -jar ${PICARD} MarkDuplicates \
$(echo ${bam_array[@]}) \
OUTPUT=${ALL_SAMPLEPATH}/{FLOWCELL}_allSamples_markdup.bam \
METRICS_FILE=${ALL_SAMPLEPATH}/dupmetrics_${FLOWCELL}_allSamples.txt \
CREATE_INDEX=true \
REMOVE_DUPLICATES=true \
MAX_RECORDS_IN_RAM=25000000 \
ASSUME_SORT_ORDER=coordinate \
USE_JDK_DEFLATER=true \
USE_JDK_INFLATER=true \
TMP_DIR=/tmp)>"markdup.log" 2>&1
wait;


#check coverage on all sample bam (should be between 20-30x)
export FLOWCELL=$(echo $(pwd) | cut -d "/" -f5);echo $FLOWCELL
echo "Plotting Coverage ${FLOWCELL}"
(plotCoverage --bamfiles ${FLOWCELL}_allSamples_markdup.bam \
--plotFile coverageplot_${FLOWCELL}_allSamples.png \
--smartLabels \
--plotTitle "Coverage of GIAB NA12878-${FLOWCELL}" \
--outRawCounts rawcounts_${FLOWCELL}_allSamples \
--plotFileFormat png \
--numberOfProcessors max)>"coverage.log" 2>&1