/tools/gatk/table_recalibration.xml

https://bitbucket.org/cistrome/cistrome-harvard/ · XML · 232 lines · 209 code · 20 blank · 3 comment · 0 complexity · f80798fcdf7fd711945ce167933e9285 MD5 · raw file 

    

  1. <tool id="gatk_table_recalibration" name="Table Recalibration" version="0.0.5">
    2. 

  2.   <description>on BAM files</description>
    3. 

  3.   <requirements>
    4. 

  4.       <requirement type="package" version="1.4">gatk</requirement>
    5. 

  5.       <requirement type="package">samtools</requirement>
    6. 

  6.   </requirements>
    7. 

  7.   <macros>
    8. 

  8.     <import>gatk_macros.xml</import>
    9. 

  9.   </macros>
    10. 

  10.   <command interpreter="python">gatk_wrapper.py
    11. 

  11.    --max_jvm_heap_fraction "1"
    12. 

  12.    --stdout "${output_log}"
    13. 

  13.    -d "-I" "${reference_source.input_bam}" "${reference_source.input_bam.ext}" "gatk_input"
    14. 

  14.    #if str( $reference_source.input_bam.metadata.bam_index ) != "None":
    15. 

  15.        -d "" "${reference_source.input_bam.metadata.bam_index}" "bam_index" "gatk_input" ##hardcode galaxy ext type as bam_index
    16. 

  16.    #end if
    17. 

  17.    -p 'java 
    18. 

  18.     -jar "${GALAXY_DATA_INDEX_DIR}/shared/jars/gatk/GenomeAnalysisTK.jar"
    19. 

  19.     -T "TableRecalibration"
    20. 

  20.     -o "${output_bam}"
    21. 

  21.     -et "NO_ET" ##ET no phone home
    22. 

  22.     ##--num_threads 4 ##hard coded, for now
    23. 

  23.     ##-log "${output_log}" ##don't use this to log to file, instead directly capture stdout
    24. 

  24.     #if $reference_source.reference_source_selector != "history":
    25. 

  25.         -R "${reference_source.ref_file.fields.path}"
    26. 

  26.     #end if
    27. 

  27.     --recal_file "${input_recal}"
    28. 

  28.     --disable_bam_indexing
    29. 

  29.    '
    30. 

  30.     #include source=$standard_gatk_options#
    31. 

  31.     
    32. 

  32.     ##start analysis specific options
    33. 

  33.     #if $analysis_param_type.analysis_param_type_selector == "advanced":
    34. 

  34.         -p '
    35. 

  35.         #if $analysis_param_type.default_read_group_type.default_read_group_type_selector == "set":
    36. 

  36.             --default_read_group "${analysis_param_type.default_read_group_type.default_read_group}"
    37. 

  37.         #end if
    38. 

  38.         #if str( $analysis_param_type.default_platform ) != "default":
    39. 

  39.             --default_platform "${analysis_param_type.default_platform}"
    40. 

  40.         #end if
    41. 

  41.         #if str( $analysis_param_type.force_read_group_type.force_read_group_type_selector ) == "set":
    42. 

  42.             --force_read_group "${analysis_param_type.force_read_group_type.force_read_group}"
    43. 

  43.         #end if
    44. 

  44.         #if str( $analysis_param_type.force_platform ) != "default":
    45. 

  45.             --force_platform "${analysis_param_type.force_platform}"
    46. 

  46.         #end if
    47. 

  47.         ${analysis_param_type.exception_if_no_tile}
    48. 

  48.         #if str( $analysis_param_type.solid_options_type.solid_options_type_selector ) == "set":
    49. 

  49.             #if str( $analysis_param_type.solid_options_type.solid_recal_mode ) != "default":
    50. 

  50.                 --solid_recal_mode "${analysis_param_type.solid_options_type.solid_recal_mode}" 
    51. 

  51.             #end if
    52. 

  52.             #if str( $analysis_param_type.solid_options_type.solid_nocall_strategy ) != "default":
    53. 

  53.                 --solid_nocall_strategy "${analysis_param_type.solid_options_type.solid_nocall_strategy}" 
    54. 

  54.             #end if
    55. 

  55.         #end if
    56. 

  56.         ${analysis_param_type.simplify_bam}
    57. 

  57.         --preserve_qscores_less_than "${analysis_param_type.preserve_qscores_less_than}"
    58. 

  58.         --smoothing "${analysis_param_type.smoothing}"
    59. 

  59.         --max_quality_score "${analysis_param_type.max_quality_score}"
    60. 

  60.         --window_size_nqs "${analysis_param_type.window_size_nqs}"
    61. 

  61.         --homopolymer_nback "${analysis_param_type.homopolymer_nback}"
    62. 

  62.         ${analysis_param_type.do_not_write_original_quals}
    63. 

  63.         '
    64. 

  64.     #end if
    65. 

  65.   </command>
    66. 

  66.   <inputs>
    67. 

  67.     <param name="input_recal" type="data" format="csv" label="Covariates table recalibration file" help="-recalFile,--recal_file &amp;lt;recal_file&amp;gt;" />
    68. 

  68.     <conditional name="reference_source">
    69. 

  69.       <expand macro="reference_source_selector_param" />
    70. 

  70.       <when value="cached">
    71. 

  71.         <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &amp;lt;input_file&amp;gt;">
    72. 

  72.           <validator type="unspecified_build" />
    73. 

  73.           <validator type="dataset_metadata_in_data_table" table_name="gatk_picard_indexes" metadata_name="dbkey" metadata_column="dbkey" message="Sequences are not currently available for the specified build." /> <!-- fixme!!! this needs to be a select -->
    74. 

  74.         </param>
    75. 

  75.         <param name="ref_file" type="select" label="Using reference genome" help="-R,--reference_sequence &amp;lt;reference_sequence&amp;gt;" >
    76. 

  76.           <options from_data_table="gatk_picard_indexes">
    77. 

  77.             <filter type="data_meta" key="dbkey" ref="input_bam" column="dbkey"/>
    78. 

  78.           </options>
    79. 

  79.           <validator type="no_options" message="A built-in reference genome is not available for the build associated with the selected input file"/>
    80. 

  80.         </param>
    81. 

  81.       </when>
    82. 

  82.       <when value="history">
    83. 

  83.         <param name="input_bam" type="data" format="bam" label="BAM file" help="-I,--input_file &amp;lt;input_file&amp;gt;" />
    84. 

  84.         <param name="ref_file" type="data" format="fasta" label="Using reference file" help="-R,--reference_sequence &amp;lt;reference_sequence&amp;gt;">
    85. 

  85.           <options>
    86. 

  86.             <filter type="data_meta" key="dbkey" ref="input_bam" />
    87. 

  87.           </options>
    88. 

  88.         </param>
    89. 

  89.       </when>
    90. 

  90.     </conditional>
    91. 

  91.     
    92. 

  92.     <expand macro="gatk_param_type_conditional" />
    93. 

  93.     
    94. 

  94.     
    95. 

  95.     <expand macro="analysis_type_conditional">
    96. 

  96.         <conditional name="default_read_group_type">
    97. 

  97.           <param name="default_read_group_type_selector" type="select" label="Set default Read Group" help="--default_read_group">
    98. 

  98.             <option value="default" selected="True">Don't Set</option>
    99. 

  99.             <option value="set">Set</option>
    100. 

  100.           </param>
    101. 

  101.           <when value="default">
    102. 

  102.             <!-- do nothing here -->
    103. 

  103.           </when>
    104. 

  104.           <when value="set">
    105. 

  105.             <param name="default_read_group" type="text" value="Unknown" label="If a read has no read group then default to the provided String"/>
    106. 

  106.           </when>
    107. 

  107.         </conditional>
    108. 

  108.         <param name="default_platform" type="select" label="Set default Platform" help="--default_platform">
    109. 

  109.           <option value="default" selected="True">Don't Set</option>
    110. 

  110.           <option value="illumina">illumina</option>
    111. 

  111.           <option value="454">454</option>
    112. 

  112.           <option value="solid">solid</option>
    113. 

  113.         </param>
    114. 

  114.         <conditional name="force_read_group_type">
    115. 

  115.           <param name="force_read_group_type_selector" type="select" label="Force Read Group" help="--force_read_group">
    116. 

  116.             <option value="default" selected="True">Don't Force</option>
    117. 

  117.             <option value="set">Force</option>
    118. 

  118.           </param>
    119. 

  119.           <when value="default">
    120. 

  120.             <!-- do nothing here -->
    121. 

  121.           </when>
    122. 

  122.           <when value="set">
    123. 

  123.             <param name="force_read_group" type="text" value="Unknown" label="If provided, the read group ID of EVERY read will be forced to be the provided String."/>
    124. 

  124.           </when>
    125. 

  125.         </conditional>
    126. 

  126.         <param name="force_platform" type="select" label="Force Platform" help="--force_platform">
    127. 

  127.           <option value="default" selected="True">Don't Force</option>
    128. 

  128.           <option value="illumina">illumina</option>
    129. 

  129.           <option value="454">454</option>
    130. 

  130.           <option value="solid">solid</option>
    131. 

  131.         </param>
    132. 

  132.         <param name="exception_if_no_tile" type="boolean" checked="False" truevalue="--exception_if_no_tile" falsevalue="" label="Throw an exception when no tile can be found" help="--exception_if_no_tile"/>
    133. 

  133.         <conditional name="solid_options_type">
    134. 

  134.           <param name="solid_options_type_selector" type="select" label="Set SOLiD specific options">
    135. 

  135.             <option value="default" selected="True">Don't Set</option>
    136. 

  136.             <option value="set">Set</option>
    137. 

  137.           </param>
    138. 

  138.           <when value="default">
    139. 

  139.             <!-- do nothing here -->
    140. 

  140.           </when>
    141. 

  141.           <when value="set">
    142. 

  142.             <param name="solid_recal_mode" type="select" label="How should we recalibrate solid bases in which the reference was inserted" help="-sMode,--solid_recal_mode &amp;lt;solid_recal_mode&amp;gt;">
    143. 

  143.               <option value="default" selected="True">Don't set</option>
    144. 

  144.               <option value="DO_NOTHING">DO_NOTHING</option>
    145. 

  145.               <option value="SET_Q_ZERO">SET_Q_ZERO</option>
    146. 

  146.               <option value="SET_Q_ZERO_BASE_N">SET_Q_ZERO_BASE_N</option>
    147. 

  147.               <option value="REMOVE_REF_BIAS">REMOVE_REF_BIAS</option>
    148. 

  148.             </param>
    149. 

  149.             <param name="solid_nocall_strategy" type="select" label="Behavior of the recalibrator when it encounters no calls" help="-solid_nocall_strategy,--solid_nocall_strategy &amp;lt;solid_nocall_strategy&amp;gt;">
    150. 

  150.               <option value="default" selected="True">Don't set</option>
    151. 

  151.               <option value="THROW_EXCEPTION">THROW_EXCEPTION</option>
    152. 

  152.               <option value="LEAVE_READ_UNRECALIBRATED">LEAVE_READ_UNRECALIBRATED</option>
    153. 

  153.               <option value="PURGE_READ">PURGE_READ</option>
    154. 

  154.             </param>
    155. 

  155.           </when>
    156. 

  156.         </conditional>
    157. 

  157.         <param name="simplify_bam" type="boolean" checked="False" truevalue="-simplifyBAM" falsevalue="" label="Simplify BAM" help="-simplifyBAM,--simplifyBAM"/>
    158. 

  158.         <param name="window_size_nqs" type="integer" value="5" label="Window size used by MinimumNQSCovariate" help="--window_size_nqs"/>
    159. 

  159.         <param name="homopolymer_nback" type="integer" value="7" label="Number of previous bases to look at in HomopolymerCovariate" help="-nback,--homopolymer_nback &amp;lt;homopolymer_nback&amp;gt;" />
    160. 

  160.         <param name="preserve_qscores_less_than" type="integer" value="5" label="Bases with quality scores less than this threshold won't be recalibrated" help="-pQ,--preserve_qscores_less_than &amp;lt;preserve_qscores_less_than&amp;gt;"/>
    161. 

  161.         <param name="smoothing" type="integer" value="1" label="smoothing" help="-sm,--smoothing &amp;lt;smoothing&amp;gt;"/>
    162. 

  162.         <param name="max_quality_score" type="integer" value="50" label="Max quality score" help="-maxQ,--max_quality_score &amp;lt;max_quality_score&amp;gt;"/>
    163. 

  163.         <param name="do_not_write_original_quals" type="boolean" checked="False" truevalue="--doNotWriteOriginalQuals" falsevalue="" label="Do Not Write Original Quality tag" help="-noOQs,--doNotWriteOriginalQuals"/>
    164. 

  164.     </expand>
    165. 

  165.   </inputs>
    166. 

  166.   <outputs>
    167. 

  167.     <data format="bam" name="output_bam" label="${tool.name} on ${on_string} (BAM)" />
    168. 

  168.     <data format="txt" name="output_log" label="${tool.name} on ${on_string} (log)" />
    169. 

  169.   </outputs>
    170. 

  170.   <tests>
    171. 

  171.       <test>
    172. 

  172.           <param name="input_recal" value="gatk/gatk_count_covariates/gatk_count_covariates_out_1.csv" ftype="csv" /> 
    173. 

  173.           <param name="reference_source_selector" value="history" />
    174. 

  174.           <param name="ref_file" value="phiX.fasta" ftype="fasta" />
    175. 

  175.           <param name="input_bam" value="gatk/gatk_indel_realigner/gatk_indel_realigner_out_1.bam" ftype="bam" />
    176. 

  176.           <param name="gatk_param_type_selector" value="basic" />
    177. 

  177.           <param name="analysis_param_type_selector" value="basic" />
    178. 

  178.           <output name="output_bam" file="gatk/gatk_table_recalibration/gatk_table_recalibration_out_1.bam" ftype="bam" lines_diff="4" />
    179. 

  179.           <output name="output_log" file="gatk/gatk_table_recalibration/gatk_table_recalibration_out_1.log.contains" compare="contains" />
    180. 

  180.       </test>
    181. 

  181.   </tests>
    182. 

  182.   <help>
    183. 

  183. **What it does**
    184. 

  184. 

  185. This walker is designed to work as the second pass in a two-pass processing step, doing a by-read traversal.  For each base in each read this walker calculates various user-specified covariates (such as read group, reported quality score, cycle, and dinuc) Using these values as a key in a large hashmap the walker calculates an empirical base quality score and overwrites the quality score currently in the read. This walker then outputs a new bam file with these updated (recalibrated) reads.  Note: This walker expects as input the recalibration table file generated previously by CovariateCounterWalker. Note: This walker is designed to be used in conjunction with CovariateCounterWalker.
    186. 

  186. 

  187. For more information on base quality score recalibration using the GATK, see this `tool specific page &lt;http://www.broadinstitute.org/gsa/wiki/index.php/Base_quality_score_recalibration&gt;`_.
    188. 

  188. 

  189. To learn about best practices for variant detection using GATK, see this `overview &lt;http://www.broadinstitute.org/gsa/wiki/index.php/Best_Practice_Variant_Detection_with_the_GATK_v3&gt;`_.
    190. 

  190. 

  191. If you encounter errors, please view the `GATK FAQ &lt;http://www.broadinstitute.org/gsa/wiki/index.php/Frequently_Asked_Questions&gt;`_.
    192. 

  192. 

  193. ------
    194. 

  194. 

  195. **Inputs**
    196. 

  196. 

  197. GenomeAnalysisTK: TableRecalibration accepts an aligned BAM and a recalibration CSV input files.
    198. 

  198. 

  199. 

  200. **Outputs**
    201. 

  201. 

  202. The output is in BAM format.
    203. 

  203. 

  204. 

  205. Go `here &lt;http://www.broadinstitute.org/gsa/wiki/index.php/Input_files_for_the_GATK&gt;`_ for details on GATK file formats.
    206. 

  206. 

  207. -------
    208. 

  208. 

  209. **Settings**::
    210. 

  210. 

  211.  default_read_group             If a read has no read group then default to the provided String.
    212. 

  212.  default_platform               If a read has no platform then default to the provided String. Valid options are illumina, 454, and solid.
    213. 

  213.  force_read_group               If provided, the read group ID of EVERY read will be forced to be the provided String. This is useful to collapse all data into a single read group.
    214. 

  214.  force_platform                 If provided, the platform of EVERY read will be forced to be the provided String. Valid options are illumina, 454, and solid.
    215. 

  215.  window_size_nqs                The window size used by MinimumNQSCovariate for its calculation
    216. 

  216.  homopolymer_nback              The number of previous bases to look at in HomopolymerCovariate
    217. 

  217.  exception_if_no_tile           If provided, TileCovariate will throw an exception when no tile can be found. The default behavior is to use tile = -1
    218. 

  218.  solid_recal_mode               How should we recalibrate solid bases in whichthe reference was inserted? Options = DO_NOTHING, SET_Q_ZERO, SET_Q_ZERO_BASE_N, or REMOVE_REF_BIAS (DO_NOTHING|SET_Q_ZERO|SET_Q_ZERO_BASE_N|REMOVE_REF_BIAS)
    219. 

  219.  solid_nocall_strategy          Defines the behavior of the recalibrator when it encounters no calls in the color space. Options = THROW_EXCEPTION, LEAVE_READ_UNRECALIBRATED, or PURGE_READ (THROW_EXCEPTION|LEAVE_READ_UNRECALIBRATED|PURGE_READ)
    220. 

  220.  recal_file                     Filename for the input covariates table recalibration .csv file
    221. 

  221.  out                            The output BAM file
    222. 

  222.  bam_compression                Compression level to use for writing BAM files
    223. 

  223.  disable_bam_indexing           Turn off on-the-fly creation of indices for output BAM files.
    224. 

  224.  simplifyBAM                    If provided, output BAM files will be simplified to include just key reads for downstream variation discovery analyses (removing duplicates, PF-, non-primary reads), as well stripping all extended tags from the kept reads except the read group identifier
    225. 

  225.  preserve_qscores_less_than     Bases with quality scores less than this threshold won't be recalibrated, default=5. In general it's unsafe to change qualities scores below &lt; 5, since base callers use these values to indicate random or bad bases
    226. 

  226.  smoothing                      Number of imaginary counts to add to each bin bin order to smooth out bins with few data points, default=1
    227. 

  227.  max_quality_score              The integer value at which to cap the quality scores, default=50
    228. 

  228.  doNotWriteOriginalQuals        If true, we will not write the original quality (OQ) tag for each read
    229. 

  229. 

  230. @CITATION_SECTION@
    231. 

  231.   </help>
    232. 

  232. </tool>
    233.