Python初体验

今天开始所有的工作脚本全都从perl转变到python，开发速度明显降低了不少，相信以后随着熟练度提升会好起来。贴一下今天一个工作代码，由于之前去一家小公司测序时，序列长度竟然都没有达到要求，为了之后的索赔事宜，写了个脚本统计所有序列的结果，主要包括总的reads数，bases数，和达到测序策略要求长度的reads数（双端），bases数，高质量（Q30）bases数，高质量reads数（双端）等等......多个文件的统计工作一般会写一个单独处理一个文件的脚本，然后再写一个脚本用来生成多个文件的处理的shell脚本，然后想办法并行处理这个shell就可以，效率会快很多。由于测序数据往往比较大，IO操作时，逐行读取是上策。

统计单个文件测序数据情况脚本：

 from __future__ import division
 from Bio import SeqIO as fq
 import os
 import sys
 import re
 read1_gzfile = sys.argv[1]
 read2_gzfile = sys.argv[2]
 gz_handle1 = os.popen( 'gunzip -cd %s' % read1_gzfile )
 gz_handle2 = os.popen( 'gunzip -cd %s' % read2_gzfile )
 basename1 = os.path.basename(read1_gzfile)
 basename1 = re.match('(\S+)_R1_001\.fastq\.gz',basename1).group(1)
 basename2 = os.path.basename(read2_gzfile)
 basename2 = re.match('(\S+)_R2_001\.fastq\.gz',basename2).group(1)
 if basename1 != basename2:
     raise 'Two Read are not mapped!'
 cwd = os.getcwd()
 out_handle = open('%s/%s.stat'%(cwd,basename1),'w')
 out_handle.write('AllReadsNum\tRead1_PE300_ReadsNum\tRead2_PE300_ReadsNum\tUseful_ReadsNum(Read1>=300 and Read2>=300)\tAll_Bases\tRead1_Q30_Bases(PE300)\tRead2_Q30_Bases(PE300)\tQ30_PE_Reads(Q30>50%)\tUseful_Bases(All)\tUseful_Ratio\n')

 AllReadsNum = 0
 AllBases = 0
 Read1_PE300_ReadsNum = 0
 Read2_PE300_ReadsNum = 0
 Useful_ReadsNum = 0
 Read1_Q30_Bases = 0
 Read2_Q30_Bases = 0
 Q30_PE_Reads = 0
 Useful_Bases = 0

 def PE300(seq):
     if len(seq) >= 300:
         return True
     else:
         return False

 def Q30(qual_list):
     num = 0
     for qual in qual_list:
         if qual >= 30:
             num += 1
     return num

 reads2 = fq.parse(gz_handle2,'fastq')
 for read1 in fq.parse(gz_handle1,'fastq'):
     read2 = reads2.next()
     seq1 = read1.seq
     qual1 = read1.letter_annotations['phred_quality']
     seq2 = read2.seq
     qual2 = read2.letter_annotations['phred_quality']
     AllReadsNum += 1
     AllBases += len(seq1)
     AllBases += len(seq2)
     R1_300 = PE300(seq1)
     R2_300 = PE300(seq2)
     if R1_300 and R2_300:
         Useful_ReadsNum +=1
         R1_Q30 = Q30(qual1)
         R2_Q30 = Q30(qual2)
         Read1_Q30_Bases += R1_Q30
         Read2_Q30_Bases += R2_Q30
         if ( R1_Q30 / len(seq1) >= 0.5 ) and ( R2_Q30 / len(seq2) >= 0.5 ):
             Q30_PE_Reads += 1
             Useful_Bases += R1_Q30
             Useful_Bases += R2_Q30
     elif R1_300:
         Read1_PE300_ReadsNum += 1
     elif R2_300:
         Read2_PE300_ReadsNum += 1

 Useful_Ratio = Useful_Bases / AllBases
 out_handle.write('%i\t%i\t%i\t%i\t%i\t%i\t%i\t%i\t%i\t%f\n'%(AllReadsNum,Read1_PE300_ReadsNum,Read2_PE300_ReadsNum,Useful_ReadsNum,AllBases,Read1_Q30_Bases,Read2_Q30_Bases,Q30_PE_Reads,Useful_Bases,Useful_Ratio))

summary.py

生成脚本：

 import os
 out = open('summary.sh','w')
 cwd = os.getcwd()
 with open('templist') as gzfiles:
     for gzfile1 in gzfiles:
         gzfile2 = gzfiles.next()
         out.write('python %s/summary.py %s %s\n'%(cwd,gzfile1.strip(),gzfile2.strip()))

run_summary.py

使用qsub_sge方法，并行投递生成的summary.sh就完成了