annotate picard_MarkDuplicates.xml @ 8:3a3234d7a2e8 draft

planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tools/picard commit 00a7926c285bc4a339bd7deebf40b28f39c7d947-dirty
author devteam
date Thu, 16 Jul 2015 15:53:10 -0400
parents 4ff1e04010e6
children 5eaa8a968300
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
8
3a3234d7a2e8 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tools/picard commit 00a7926c285bc4a339bd7deebf40b28f39c7d947-dirty
devteam
parents: 6
diff changeset
1 <tool name="MarkDuplicates" id="picard_MarkDuplicates" version="@TOOL_VERSION@.0">
5
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
2 <description>examine aligned records in BAM datasets to locate duplicate molecules</description>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
3 <macros>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
4 <import>picard_macros.xml</import>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
5 </macros>
8
3a3234d7a2e8 planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tools/picard commit 00a7926c285bc4a339bd7deebf40b28f39c7d947-dirty
devteam
parents: 6
diff changeset
6 <expand macro="requirements" />
5
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
7 <command>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
8 @java_options@
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
9
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
10 java -jar \$JAVA_JAR_PATH/picard.jar
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
11 MarkDuplicates
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
12
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
13 INPUT="${inputFile}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
14 OUTPUT="${outFile}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
15
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
16 METRICS_FILE="${metrics_file}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
17 #for $element in $comments:
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
18 COMMENT="${element.comment}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
19 #end for
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
20 REMOVE_DUPLICATES="${remove_duplicates}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
21 ASSUME_SORTED="${assume_sorted}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
22
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
23 DUPLICATE_SCORING_STRATEGY="${duplicate_scoring_strategy}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
24
6
4ff1e04010e6 Uploaded
devteam
parents: 5
diff changeset
25 #import pipes
4ff1e04010e6 Uploaded
devteam
parents: 5
diff changeset
26 READ_NAME_REGEX=${ pipes.quote( str( $read_name_regex ) ) or "''" }
5
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
27 OPTICAL_DUPLICATE_PIXEL_DISTANCE="${optical_duplicate_pixel_distance}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
28
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
29 VALIDATION_STRINGENCY="${validation_stringency}"
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
30 QUIET=true
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
31 VERBOSITY=ERROR
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
32
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
33 </command>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
34 <inputs>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
35 <param format="bam" name="inputFile" type="data" label="Select SAM/BAM dataset or dataset collection" help="If empty, upload or import a SAM/BAM dataset"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
36 <repeat name="comments" title="Comment" min="0" help="You can provide multiple comments">
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
37 <param name="comment" type="text" size="50" label="Add this comment to BAM dataset"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
38 </repeat>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
39 <param name="remove_duplicates" type="boolean" label="If true do not write duplicates to the output file instead of writing them with appropriate flags set" help="REMOVE_DUPLICATES; default=False"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
40 <param name="assume_sorted" type="boolean" label="Assume the input file is already sorted" checked="true" truevalue="true" falsevalue="false" help="ASSUME_SORTED; default=True"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
41
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
42 <param name="duplicate_scoring_strategy" type="select" label="The scoring strategy for choosing the non-duplicate among candidates" help="DUPLICATE_SCORING_STRATEGY; default=SUM_OF_BASE_QUALITIES">
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
43 <option value="SUM_OF_BASE_QUALITIES">SUM_OF_BASE_QUALITIES</option>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
44 <option value="TOTAL_MAPPED_REFERENCE_LENGTH">TOTAL_MAPPED_REFERENCE_LENGTH</option>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
45 </param>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
46
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
47
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
48 <param name="read_name_regex" type="text" size="40" value="[a-zA-Z0-9]+:[0-9]:([0-9]+):([0-9]+):([0-9]+).*." label="Regular expression that can be used to parse read names in the incoming SAM/BAM dataset" help="READ_NAME_REGEX; Read names are parsed to extract three variables: tile/region, x coordinate and y coordinate. These values are used to estimate the rate of optical duplication in order to give a more accurate estimated library size. See help below for more info; default=[a-zA-Z0-9]+:[0-9]:([0-9]+):([0-9]+):([0-9]+).*.">
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
49 <sanitizer>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
50 <valid initial="string.printable">
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
51 </valid>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
52 </sanitizer>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
53 </param>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
54 <param name="optical_duplicate_pixel_distance" type="integer" value="100" min="0" max="500" label="The maximum offset between two duplicte clusters in order to consider them optical duplicates" help="OPTICAL_DUPLICATE_PIXEL_DISTANCE; default=100"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
55
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
56 <expand macro="VS" />
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
57
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
58 </inputs>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
59
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
60 <outputs>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
61 <data format="txt" name="metrics_file" label="${tool.name} on ${on_string}: MarkDuplicate metrics"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
62 <data format="bam" name="outFile" label="${tool.name} on ${on_string}: MarkDuplicates BAM output"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
63 </outputs>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
64
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
65 <tests>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
66 <test>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
67 <param name="inputFile" value="picard_MarkDuplicates.bam" ftype="bam"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
68 <param name="comment" value="test-run"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
69 <param name="assume_sorted" value="True"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
70 <param name="remove_duplicates" value="True"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
71 <param name="read_name_regex" value="[a-zA-Z0-9]+:[0-9]:([0-9]+):([0-9]+):([0-9]+).*."/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
72 <param name="optical_duplicate_pixel_distance" value="100"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
73 <param name="duplicate_scoring_strategy" value="SUM_OF_BASE_QUALITIES"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
74 <param name="validation_stringency" value="LENIENT"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
75 <output name="outFile" file="picard_MarkDuplicates_test1.bam" ftype="bam" lines_diff="2"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
76 </test>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
77 </tests>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
78
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
79 <stdio>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
80 <exit_code range="1:" level="fatal"/>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
81 </stdio>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
82
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
83 <help>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
84
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
85 **Purpose**
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
86
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
87 Examines aligned records in the supplied SAM or BAM dataset to locate duplicate molecules. All records are then written to the output file with the duplicate records flagged.
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
88
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
89 @dataset_collections@
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
90
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
91 @description@
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
92
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
93 COMMENT=String
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
94 CO=String Comment(s) to include in the output file's header. This option may be specified 0 or
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
95 more times.
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
96
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
97 REMOVE_DUPLICATES=Boolean If true do not write duplicates to the output file instead of writing them with
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
98 appropriate flags set. Default value: false.
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
99
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
100 READ_NAME_REGEX=String Regular expression that can be used to parse read names in the incoming SAM file. Read
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
101 names are parsed to extract three variables: tile/region, x coordinate and y coordinate.
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
102 These values are used to estimate the rate of optical duplication in order to give a more
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
103 accurate estimated library size. Set this option to null to disable optical duplicate
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
104 detection. The regular expression should contain three capture groups for the three
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
105 variables, in order. It must match the entire read name. Note that if the default regex
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
106 is specified, a regex match is not actually done, but instead the read name is split on
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
107 colon character. For 5 element names, the 3rd, 4th and 5th elements are assumed to be
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
108 tile, x and y values. For 7 element names (CASAVA 1.8), the 5th, 6th, and 7th elements
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
109 are assumed to be tile, x and y values. Default value:
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
110 [a-zA-Z0-9]+:[0-9]:([0-9]+):([0-9]+):([0-9]+).*.
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
111
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
112 DUPLICATE_SCORING_STRATEGY=ScoringStrategy
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
113 DS=ScoringStrategy The scoring strategy for choosing the non-duplicate among candidates. Default value:
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
114 SUM_OF_BASE_QUALITIES. Possible values: {SUM_OF_BASE_QUALITIES, TOTAL_MAPPED_REFERENCE_LENGTH}
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
115
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
116 OPTICAL_DUPLICATE_PIXEL_DISTANCE=Integer
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
117 The maximum offset between two duplicte clusters in order to consider them optical
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
118 duplicates. This should usually be set to some fairly small number (e.g. 5-10 pixels)
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
119 unless using later versions of the Illumina pipeline that multiply pixel values by 10, in
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
120 which case 50-100 is more normal. Default value: 100.
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
121
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
122 @more_info@
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
123
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
124 </help>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
125 </tool>
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
126
3d4f1fa26f0e Uploaded
devteam
parents:
diff changeset
127