reads on various attributes fastq_manipulation.py $input_file $fastq_manipulation_file $output_file $output_file.files_path '${input_file.extension[len( 'fastq' ):]}' Name/Identifier Sequence Content Quality Score Content Regular Expression Regular Expression Regular Expression Name/Identifier Sequence Content Quality Score Content Miscellaneous Actions String Translate Reverse Complement Reverse, No Complement Complement, No Reverse Trim DNA to RNA RNA to DNA String Translate Change Adapter Base Absolute Values Percentage of Read Length int( float( value ) ) == float( value ) int( float( value ) ) == float( value ) String Translate Remove Read ##create an importable module #import binascii import re import binascii from string import maketrans ##does read match def match_read( fastq_read ): #for $match_block in $match_blocks: #if $match_block['match_type']['match_type_selector'] == 'identifier': search_target = fastq_read.identifier[1:] ##don't include @ #elif $match_block['match_type']['match_type_selector'] == 'sequence': search_target = fastq_read.sequence #elif $match_block['match_type']['match_type_selector'] == 'quality': search_target = fastq_read.quality #else: #continue #end if if not re.search( binascii.unhexlify( "${ binascii.hexlify( str( match_block['match_type']['match']['match_by'] ) ) }" ), search_target ): return False #end for return True ##modify matched reads def manipulate_read( fastq_read ): new_read = fastq_read.clone() #for $manipulate_block in $manipulate_blocks: #if $manipulate_block['manipulation_type']['manipulation_type_selector'] == 'identifier': #if $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'translate': new_read.identifier = "@%s" % new_read.identifier[1:].translate( maketrans( binascii.unhexlify( "${ binascii.hexlify( str( manipulate_block['manipulation_type']['manipulation']['from'] ) ) }" ), binascii.unhexlify( "${ binascii.hexlify( str( manipulate_block['manipulation_type']['manipulation']['to'] ) ) }" ) ) ) #end if #elif $manipulate_block['manipulation_type']['manipulation_type_selector'] == 'sequence': #if $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'translate': new_read.sequence = new_read.sequence.translate( maketrans( binascii.unhexlify( "${ binascii.hexlify( str( manipulate_block['manipulation_type']['manipulation']['from'] ) ) }" ), binascii.unhexlify( "${ binascii.hexlify( str( manipulate_block['manipulation_type']['manipulation']['to'] ) ) }" ) ) ) #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'rev_comp': new_read = new_read.reverse_complement() #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'rev_no_comp': new_read = new_read.reverse() #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'no_rev_comp': new_read = new_read.complement() #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'trim': #if $manipulate_block['manipulation_type']['manipulation']['offset_type']['base_offset_type'] == 'offsets_percent': left_column_offset = int( round( float( ${ manipulate_block['manipulation_type']['manipulation']['offset_type']['left_column_offset'] } ) / 100.0 * float( len( new_read ) ) ) ) right_column_offset = int( round( float( ${ manipulate_block['manipulation_type']['manipulation']['offset_type']['right_column_offset'] } ) / 100.0 * float( len( new_read ) ) ) ) #else left_column_offset = ${ manipulate_block['manipulation_type']['manipulation']['offset_type']['left_column_offset'] } right_column_offset = ${ manipulate_block['manipulation_type']['manipulation']['offset_type']['right_column_offset'] } #end if if right_column_offset > 0: right_column_offset = -right_column_offset else: right_column_offset = None new_read = new_read.slice( left_column_offset, right_column_offset ) if not ( ${str( manipulate_block['manipulation_type']['manipulation']['keep_zero_length'] ) == 'keep_zero_length'} or len( new_read ) ): return None #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'dna_to_rna': new_read = new_read.sequence_as_DNA() #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'rna_to_dna': new_read = new_read.sequence_as_RNA() #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'change_adapter': if new_read.sequence_space == 'color': new_read = new_read.change_adapter( binascii.unhexlify( "${ binascii.hexlify( str( manipulate_block['manipulation_type']['manipulation']['new_adapter'] ) ) }" ) ) #end if #elif $manipulate_block['manipulation_type']['manipulation_type_selector'] == 'quality': #if $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'translate': new_read.quality = new_read.quality.translate( maketrans( binascii.unhexlify( "${ binascii.hexlify( str( manipulate_block['manipulation_type']['manipulation']['from'] ) ) }" ), binascii.unhexlify( "${ binascii.hexlify( str( manipulate_block['manipulation_type']['manipulation']['to'] ) ) }" ) ) ) #elif $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'map_score': def score_method( score ): raise Exception, "Unimplemented" ##This option is not yet available, need to abstract out e.g. column adding tool action: preventing users from using 'harmful' actions new_read.quality_map( score_method ) #end if #elif $manipulate_block['manipulation_type']['manipulation_type_selector'] == 'miscellaneous': #if $manipulate_block['manipulation_type']['manipulation']['manipulation_selector'] == 'remove': return None #end if #else: #continue #end if #end for if new_read.description != "+": new_read.description = "+%s" % new_read.identifier[1:] ##ensure description is still valid return new_read def match_and_manipulate_read( fastq_read ): new_read = fastq_read if match_read( fastq_read ): new_read = manipulate_read( fastq_read ) return new_read

This tool allows you to build complex manipulations to be applied to each matching read in a FASTQ file. A read must match all matching directives in order for it to be manipulated; if a read does not match, it is output in a non-modified manner. All reads matching will have each of the specified manipulations performed upon them, in the order specified. Regular Expression Matches are made using re.search, see http://docs.python.org/library/re.html for more information. All matching is performed on a single line string, regardless if e.g. the sequence or quality score spans multiple lines in the original file. String translations are performed using string.translate, see http://docs.python.org/library/string.html#string.translate and http://docs.python.org/library/string.html#string.maketrans for more information. .. class:: warningmark Only color space reads can have adapter bases substituted. ----- **Example** Suppose you have a color space sanger formatted sequence (fastqcssanger) and you want to double-encode the color space into psuedo-nucleotide space (this is different from converting) to allow these reads to be used in tools which do not natively support it (using specially designed indexes). This tool can handle this manipulation, however, this is generally not recommended as results tend to be poorer than those produced from tools which are specially designed to handle color space data. Steps: 1. Click **Add new Match Reads** and leave the matching options set to the default (Matching by sequence name/identifier using the regular expression "\*."; thereby matching all reads). 2. Click **Add new Manipulate Reads**, change **Manipulate Reads on** to "Sequence Content", set **Sequence Manipulation Type** to "Change Adapter Base" and set **New Adapter** to "" (an empty text field). 3. Click **Add new Manipulate Reads**, change **Manipulate Reads on** to "Sequence Content", set **Sequence Manipulation Type** to "String Translate" and set **From** to "0123." and **To** to "ACGTN". 4. Click Execute. The new history item will contained double-encoded psuedo-nucleotide space reads.