Serial version of GC_calc.py for speedup testing purposes

serial_gc.py

def FASTA(filename):
  try:
    f = file(filename)
  except IOError:                     
    print "The file, %s, does not exist" % filename
    return

  
  sequences = {}
    
  for line in f:
    if line.startswith('>'):
      name = line[1:].rstrip('\n')
      name = name.replace('_', ' ')

      sequences[name] = ''
    else:
      sequences[name] += line.rstrip('\n').rstrip('*')

  return sequences

def getGC(sequence):
  seqLen = len(sequence)
  G = sequence.count('G')
  C = sequence.count('C')
  GC_counts = G + C
  GC_content = float(GC_counts)/float(seqLen)*100
  return GC_content


test = FASTA(sys.argv[1])

for key,value in test.iteritems():
  print (key,getGC(value) )