Mercurial > repos > devteam > fastq_paired_end_splitter
diff fastq_paired_end_splitter.py @ 0:5943afe0f24d draft
Imported from capsule None
author | devteam |
---|---|
date | Thu, 23 Jan 2014 12:32:16 -0500 |
parents | |
children |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/fastq_paired_end_splitter.py Thu Jan 23 12:32:16 2014 -0500 @@ -0,0 +1,33 @@ +#Dan Blankenberg +import sys, os, shutil +from galaxy_utils.sequence.fastq import fastqReader, fastqWriter, fastqSplitter + +def main(): + #Read command line arguments + input_filename = sys.argv[1] + input_type = sys.argv[2] or 'sanger' + output1_filename = sys.argv[3] + output2_filename = sys.argv[4] + + splitter = fastqSplitter() + out1 = fastqWriter( open( output1_filename, 'wb' ), format = input_type ) + out2 = fastqWriter( open( output2_filename, 'wb' ), format = input_type ) + + i = None + skip_count = 0 + for i, fastq_read in enumerate( fastqReader( open( input_filename, 'rb' ), format = input_type ) ): + read1, read2 = splitter.split( fastq_read ) + if read1 and read2: + out1.write( read1 ) + out2.write( read2 ) + else: + skip_count += 1 + out1.close() + out2.close() + if i is None: + print "Your file contains no valid FASTQ reads." + else: + print 'Split %s of %s reads (%.2f%%).' % ( i - skip_count + 1, i + 1, float( i - skip_count + 1 ) / float( i + 1 ) * 100.0 ) + +if __name__ == "__main__": + main()