get the mean base composition for a fastq file after running fastq_stats.py from Galaxy

fastq_summary.py

import numpy as np
from pylab import *
import sys
file=sys.argv[1]
fh=open(file,'r')
headerline=fh.readline()
As=[]
Cs=[]
Gs=[]
Ts=[]
Ns=[]

mtrx=[] # this will be a list of numpy arrays


for line in fh:
    fields=line.strip().split('\t')
    #get the base counts of A G T C Ns from the cycle number
    basecounts= np.array( [int(x) for x in fields[13:18] ], dtype=float )
    rowsum=np.sum(basecounts)
    #get the % base composition
    basecomp=np.divide(basecounts, rowsum)
    #now append it to the list of numpy arrays
    mtrx.append(basecomp)

#convert it to a proper numpy matrix
basecomptable=np.matrix(mtrx)

#get the mean percentage of each type of base across all cycles (where #cyclces equals read length)
basecompmeans=np.mean(basecomptable, axis=0)
row=basecompmeans[0,:].tolist()
for r in row:
    outstr="\t".join(map(str,r))
    print "\t".join([file, outstr])