去除测序reads中的接头：adaptor

之前用c写过一个程序，查找reads中是否包含了adaptor，如果检测到的话就过滤掉含有adaptor的reads，这次在过滤完数据之后发现接头序列比较多，为了提升组装效果，又不能很大地影响数据量，需要对接头进行截断处理，并过滤过短的reads，用python写了一个简短的程序，指定超过3个错配以内的匹配都认为匹配到，并且长度小于50bp的reads过滤，在以下程序基础上添加传入参数，可以适用比较多的情况（单端、双端、含有single等）：

 import sys
 import re
 from Bio import SeqIO

 def rmPE(read1,read2,adaptor1,adaptor2,min_length):
     res_1 = rmSE(read1,adaptor1,min_length)
     res_2 = rmSE(read2,adaptor2,min_length)
     if res_1 and res_2:
         return res_1,res_2
     else:
         return False

 def rmSE(read,adaptor,min_length):
     seq = read.seq
     seed_len = 6
     a_len = len(adaptor)
     seq_len = len(seq)
     for i in range(a_len - seed_len):
         seed = adaptor[i:i+seed_len]
         pos = 0
         while(pos < seq_len):
             find_pos = seq.find(seed,pos)
             if find_pos > 0:
                 mistaken_count = 0
                 _b = find_pos
                 _e = find_pos + seed_len
                 while(_b >= 0 and i >= find_pos - _b):
                     if adaptor[i - find_pos + _b] != seq[_b]:
                         mistaken_count += 1
                     if mistaken_count > 3:
                         break
                     _b -= 1
                 else :
                     while(_e < seq_len and i - find_pos + _e < a_len):
                         if adaptor[ i - find_pos + _e ] != seq[_e]:
                             mistaken_count += 1
                         if mistaken_count > 3:
                             break
                         _e += 1
                     else:
                         if find_pos - i > min_length:
                             return  read[:find_pos-i]
                         else :
                             return False
                 pos = find_pos + 1
             else:
                 break
     return read

 def rmAdaptor(argv):
     argv.pop(0)
     type = argv.pop(0)
     if type=='PE':
         read1_file,read2_file,adaptor1,adaptor2,out_prefix,min_length = argv
         read2_records = SeqIO.parse(open(read2_file),'fastq')
         read1_out = open( '%s.1.fq'%out_prefix,'w' )
         read2_out = open( '%s.2.fq'%out_prefix,'w' )
         for read1 in SeqIO.parse(open(read1_file),'fastq'):
             read2 = read2_records.next()
             rmPE_res = rmPE(read1,read2,adaptor1,adaptor2,min_length)
             if rmPE_res:
                 read1_out.write(rmPE_res[0].format('fastq'))
                 read2_out.write(rmPE_res[1].format('fastq'))
     elif type=='SE':
         reads_file,adaptor,out_prefix,min_length = argv
         reads_out = open( '%s.single.fq'%out_prefix,'w' )
         for reads in SeqIO.parse(open(reads_file),'fastq'):
             rmSE_res = False
             if re.search('[\s\/](\d)',reads.id).group(1) == '':
                 rmSE_res = rmSE(reads,adaptor1,min_length)
             elif re.search('[\s\/](\d)',reads.id).group(1) == '':
                 rmSE_res = rmSE(reads,adaptor2,min_length)
             if rmSE_res:
                 reads_out.write(rmSE_res.format('fastq'))

 if __name__ == '__main__':
     rmAdaptor(sys.argv)