java -jar cromwell.jar run ercc.wdl --inputs inputs.json

   zcat FASTQ_R1 | paste - - - - | wc -l  
 

   bwa mem -t THREADS -M REF_GENOME FASTQ_R1 FASTQ_R2 | samtools view -S - | 
       cut -f 3 | grep ERCC | sort | uniq -c | sed s/^\ *// > CNV_FILE

  python <<CODE
  import os
  ercc = os.path.expandvars("~{erccData}")
  counts = "~{erccCounts}"
  total = ~{totalReads}
  output_file = "~{basename(erccCounts, '.csv')}_rpkm.csv"
 
  # 1. Open file, read and collect length data
  ercc_data = open(ercc, 'r')
  length_hash = {}
  for nextLine in ercc_data.readlines():
      nextFields = nextLine.split("\t")
      length_hash[nextFields[0]] = int(nextFields[4])
  ercc_data.close()
 
  # 2. Calculate RPKMs, put them in a hash
  count_data = open(counts, 'r')
  rpkm_hash = {}
  count_hash = {}
 
  for nextLine in count_data.readlines():
      nextLine = nextLine.strip("\n")
      countChunk = nextLine.split(" ")
      if len(countChunk) == 2 and countChunk[1] in length_hash.keys():
          count_hash[countChunk[1]] = int(countChunk[0])
 
  pm = int(total)/1000000.0
  for e in sorted(length_hash.keys()):
     rpkm = 0.0
     if e in count_hash.keys():
         rpkm = count_hash[e]/pm/(length_hash[e]/1000.0)
     rpkm_hash[e] = "{0:.2f}".format(rpkm)
  count_data.close()
 
  # 3. print out RPKMS sorted by
  f = open(output_file, "w+")
  for e in sorted(rpkm_hash.keys()):
      f.write('\t'.join([e, rpkm_hash[e]]) + '\n')
  f.close()
  CODE
 

  Rscript IMAGING_SCRIPT RPKM_TABLE CONTROL_DATA ERCC_DATA PREFIX DoseResponse SAMPLES