From 31b7ebfc67677a16d8b5a042f2f3f95abca3bbac Mon Sep 17 00:00:00 2001
From: Peter van 't Hof <p.j.van_t_hof@lumc.nl>
Date: Wed, 27 Apr 2016 11:05:33 +0200
Subject: [PATCH] Changed baserecalibrator to biopet config values
---
.../gatk/broad/BaseRecalibrator.scala | 68 +++++++------------
1 file changed, 26 insertions(+), 42 deletions(-)
diff --git a/protected/biopet-gatk-extensions/src/main/scala/nl/lumc/sasc/biopet/extensions/gatk/broad/BaseRecalibrator.scala b/protected/biopet-gatk-extensions/src/main/scala/nl/lumc/sasc/biopet/extensions/gatk/broad/BaseRecalibrator.scala
index 808085092..0408b8b79 100644
--- a/protected/biopet-gatk-extensions/src/main/scala/nl/lumc/sasc/biopet/extensions/gatk/broad/BaseRecalibrator.scala
+++ b/protected/biopet-gatk-extensions/src/main/scala/nl/lumc/sasc/biopet/extensions/gatk/broad/BaseRecalibrator.scala
@@ -5,17 +5,6 @@
*/
package nl.lumc.sasc.biopet.extensions.gatk.broad
-//import java.io.File
-//
-//import nl.lumc.sasc.biopet.utils.config.Configurable
-//
-//class BaseRecalibrator(val root: Configurable) extends org.broadinstitute.gatk.queue.extensions.gatk.BaseRecalibrator with GatkGeneral {
-// if (config.contains("scattercount")) scatterCount = config("scattercount", default = 1)
-// if (config.contains("dbsnp")) knownSites :+= new File(config("dbsnp").asString)
-// if (config.contains("known_sites")) knownSites :+= new File(config("known_sites").asString)
-//}
-//
-
import java.io.File
import nl.lumc.sasc.biopet.utils.config.Configurable
@@ -32,7 +21,7 @@ class BaseRecalibrator(val root: Configurable) extends CommandLineGATK /* with S
/** A database of known polymorphic sites */
@Input(fullName = "knownSites", shortName = "knownSites", doc = "A database of known polymorphic sites", required = false, exclusiveOf = "", validation = "")
- var knownSites: Seq[File] = Nil
+ var knownSites: List[File] = config("known_sites")
/** Dependencies on any indexes of knownSites */
@Input(fullName = "knownSitesIndexes", shortName = "", doc = "Dependencies on any indexes of knownSites", required = false, exclusiveOf = "", validation = "")
@@ -43,81 +32,77 @@ class BaseRecalibrator(val root: Configurable) extends CommandLineGATK /* with S
//@Gather(classOf[org.broadinstitute.gatk.engine.recalibration.BQSRGatherer])
var out: File = _
- /** List the available covariates and exit */
- @Argument(fullName = "list", shortName = "ls", doc = "List the available covariates and exit", required = false, exclusiveOf = "", validation = "")
- var list: Boolean = _
-
/** One or more covariates to be used in the recalibration. Can be specified multiple times */
@Argument(fullName = "covariate", shortName = "cov", doc = "One or more covariates to be used in the recalibration. Can be specified multiple times", required = false, exclusiveOf = "", validation = "")
- var covariate: Seq[String] = Nil
+ var covariate: List[String] = config("covariate", default = Nil)
/** Do not use the standard set of covariates, but rather just the ones listed using the -cov argument */
@Argument(fullName = "no_standard_covs", shortName = "noStandard", doc = "Do not use the standard set of covariates, but rather just the ones listed using the -cov argument", required = false, exclusiveOf = "", validation = "")
- var no_standard_covs: Boolean = _
+ var no_standard_covs: Boolean = config("no_standard_covs", default = false)
/** If specified, allows the recalibrator to be used without a dbsnp rod. Very unsafe and for expert users only. */
@Argument(fullName = "run_without_dbsnp_potentially_ruining_quality", shortName = "run_without_dbsnp_potentially_ruining_quality", doc = "If specified, allows the recalibrator to be used without a dbsnp rod. Very unsafe and for expert users only.", required = false, exclusiveOf = "", validation = "")
- var run_without_dbsnp_potentially_ruining_quality: Boolean = _
+ var run_without_dbsnp_potentially_ruining_quality: Boolean = config("run_without_dbsnp_potentially_ruining_quality", default = false)
/** How should we recalibrate solid bases in which the reference was inserted? Options = DO_NOTHING, SET_Q_ZERO, SET_Q_ZERO_BASE_N, or REMOVE_REF_BIAS */
@Argument(fullName = "solid_recal_mode", shortName = "sMode", doc = "How should we recalibrate solid bases in which the reference was inserted? Options = DO_NOTHING, SET_Q_ZERO, SET_Q_ZERO_BASE_N, or REMOVE_REF_BIAS", required = false, exclusiveOf = "", validation = "")
- var solid_recal_mode: String = _
+ var solid_recal_mode: Option[String] = config("solid_recal_mode")
/** Defines the behavior of the recalibrator when it encounters no calls in the color space. Options = THROW_EXCEPTION, LEAVE_READ_UNRECALIBRATED, or PURGE_READ */
@Argument(fullName = "solid_nocall_strategy", shortName = "solid_nocall_strategy", doc = "Defines the behavior of the recalibrator when it encounters no calls in the color space. Options = THROW_EXCEPTION, LEAVE_READ_UNRECALIBRATED, or PURGE_READ", required = false, exclusiveOf = "", validation = "")
- var solid_nocall_strategy: String = _
+ var solid_nocall_strategy: Option[String] = config("solid_nocall_strategy")
/** Size of the k-mer context to be used for base mismatches */
@Argument(fullName = "mismatches_context_size", shortName = "mcs", doc = "Size of the k-mer context to be used for base mismatches", required = false, exclusiveOf = "", validation = "")
- var mismatches_context_size: Option[Int] = None
+ var mismatches_context_size: Option[Int] = config("mismatches_context_size")
/** Size of the k-mer context to be used for base insertions and deletions */
@Argument(fullName = "indels_context_size", shortName = "ics", doc = "Size of the k-mer context to be used for base insertions and deletions", required = false, exclusiveOf = "", validation = "")
- var indels_context_size: Option[Int] = None
+ var indels_context_size: Option[Int] = config("indels_context_size")
/** The maximum cycle value permitted for the Cycle covariate */
@Argument(fullName = "maximum_cycle_value", shortName = "maxCycle", doc = "The maximum cycle value permitted for the Cycle covariate", required = false, exclusiveOf = "", validation = "")
- var maximum_cycle_value: Option[Int] = None
+ var maximum_cycle_value: Option[Int] = config("maximum_cycle_value")
/** default quality for the base mismatches covariate */
@Argument(fullName = "mismatches_default_quality", shortName = "mdq", doc = "default quality for the base mismatches covariate", required = false, exclusiveOf = "", validation = "")
- var mismatches_default_quality: Option[Byte] = None
+ var mismatches_default_quality: Option[String] = config("mismatches_default_quality")
/** default quality for the base insertions covariate */
@Argument(fullName = "insertions_default_quality", shortName = "idq", doc = "default quality for the base insertions covariate", required = false, exclusiveOf = "", validation = "")
- var insertions_default_quality: Option[Byte] = None
+ var insertions_default_quality: Option[String] = config("insertions_default_quality")
/** default quality for the base deletions covariate */
@Argument(fullName = "deletions_default_quality", shortName = "ddq", doc = "default quality for the base deletions covariate", required = false, exclusiveOf = "", validation = "")
- var deletions_default_quality: Option[Byte] = None
+ var deletions_default_quality: Option[String] = config("deletions_default_quality")
/** minimum quality for the bases in the tail of the reads to be considered */
@Argument(fullName = "low_quality_tail", shortName = "lqt", doc = "minimum quality for the bases in the tail of the reads to be considered", required = false, exclusiveOf = "", validation = "")
- var low_quality_tail: Option[Byte] = None
+ var low_quality_tail: Option[String] = config("low_quality_tail")
/** number of distinct quality scores in the quantized output */
@Argument(fullName = "quantizing_levels", shortName = "ql", doc = "number of distinct quality scores in the quantized output", required = false, exclusiveOf = "", validation = "")
- var quantizing_levels: Option[Int] = None
+ var quantizing_levels: Option[Int] = config("quantizing_levels")
/** the binary tag covariate name if using it */
@Argument(fullName = "binary_tag_name", shortName = "bintag", doc = "the binary tag covariate name if using it", required = false, exclusiveOf = "", validation = "")
- var binary_tag_name: String = _
+ var binary_tag_name: Option[String] = config("binary_tag_name")
/** Sort the rows in the tables of reports */
@Argument(fullName = "sort_by_all_columns", shortName = "sortAllCols", doc = "Sort the rows in the tables of reports", required = false, exclusiveOf = "", validation = "")
- var sort_by_all_columns: Boolean = _
+ var sort_by_all_columns: Boolean = config("sort_by_all_columns", default = false)
/** If a read has no platform then default to the provided String. Valid options are illumina, 454, and solid. */
@Argument(fullName = "default_platform", shortName = "dP", doc = "If a read has no platform then default to the provided String. Valid options are illumina, 454, and solid.", required = false, exclusiveOf = "", validation = "")
- var default_platform: String = _
+ var default_platform: Option[String] = config("default_platform")
/** If provided, the platform of EVERY read will be forced to be the provided String. Valid options are illumina, 454, and solid. */
@Argument(fullName = "force_platform", shortName = "fP", doc = "If provided, the platform of EVERY read will be forced to be the provided String. Valid options are illumina, 454, and solid.", required = false, exclusiveOf = "", validation = "")
- var force_platform: String = _
+ var force_platform: Option[String] = config("force_platform")
/** If provided, the read group of EVERY read will be forced to be the provided String. */
@Argument(fullName = "force_readgroup", shortName = "fRG", doc = "If provided, the read group of EVERY read will be forced to be the provided String.", required = false, exclusiveOf = "", validation = "")
- var force_readgroup: String = _
+ var force_readgroup: Option[String] = config("force_readgroup")
/** If provided, log all updates to the recalibration tables to the given file. For debugging/testing purposes only */
@Output(fullName = "recal_table_update_log", shortName = "recal_table_update_log", doc = "If provided, log all updates to the recalibration tables to the given file. For debugging/testing purposes only", required = false, exclusiveOf = "", validation = "")
@@ -126,19 +111,19 @@ class BaseRecalibrator(val root: Configurable) extends CommandLineGATK /* with S
/** Max size of the k-mer context to be used for repeat covariates */
@Argument(fullName = "max_str_unit_length", shortName = "maxstr", doc = "Max size of the k-mer context to be used for repeat covariates", required = false, exclusiveOf = "", validation = "")
- var max_str_unit_length: Option[Int] = None
+ var max_str_unit_length: Option[Int] = config("max_str_unit_length")
/** Max number of repetitions to be used for repeat covariates */
@Argument(fullName = "max_repeat_length", shortName = "maxrep", doc = "Max number of repetitions to be used for repeat covariates", required = false, exclusiveOf = "", validation = "")
- var max_repeat_length: Option[Int] = None
+ var max_repeat_length: Option[Int] = config("max_repeat_length")
/** Reduce memory usage in multi-threaded code at the expense of threading efficiency */
@Argument(fullName = "lowMemoryMode", shortName = "lowMemoryMode", doc = "Reduce memory usage in multi-threaded code at the expense of threading efficiency", required = false, exclusiveOf = "", validation = "")
- var lowMemoryMode: Boolean = _
+ var lowMemoryMode: Boolean = config("lowMemoryMode", default = false)
/** BQSR BAQ gap open penalty (Phred Scaled). Default value is 40. 30 is perhaps better for whole genome call sets */
@Argument(fullName = "bqsrBAQGapOpenPenalty", shortName = "bqsrBAQGOP", doc = "BQSR BAQ gap open penalty (Phred Scaled). Default value is 40. 30 is perhaps better for whole genome call sets", required = false, exclusiveOf = "", validation = "")
- var bqsrBAQGapOpenPenalty: Option[Double] = None
+ var bqsrBAQGapOpenPenalty: Option[Double] = config("bqsrBAQGapOpenPenalty")
/** Format string for bqsrBAQGapOpenPenalty */
@Argument(fullName = "bqsrBAQGapOpenPenaltyFormat", shortName = "", doc = "Format string for bqsrBAQGapOpenPenalty", required = false, exclusiveOf = "", validation = "")
@@ -146,15 +131,15 @@ class BaseRecalibrator(val root: Configurable) extends CommandLineGATK /* with S
/** Filter out reads with CIGAR containing the N operator, instead of failing with an error */
@Argument(fullName = "filter_reads_with_N_cigar", shortName = "filterRNC", doc = "Filter out reads with CIGAR containing the N operator, instead of failing with an error", required = false, exclusiveOf = "", validation = "")
- var filter_reads_with_N_cigar: Boolean = _
+ var filter_reads_with_N_cigar: Boolean = config("filter_reads_with_N_cigar", default = false)
/** Filter out reads with mismatching numbers of bases and base qualities, instead of failing with an error */
@Argument(fullName = "filter_mismatching_base_and_quals", shortName = "filterMBQ", doc = "Filter out reads with mismatching numbers of bases and base qualities, instead of failing with an error", required = false, exclusiveOf = "", validation = "")
- var filter_mismatching_base_and_quals: Boolean = _
+ var filter_mismatching_base_and_quals: Boolean = config("filter_mismatching_base_and_quals", default = false)
/** Filter out reads with no stored bases (i.e. '*' where the sequence should be), instead of failing with an error */
@Argument(fullName = "filter_bases_not_stored", shortName = "filterNoBases", doc = "Filter out reads with no stored bases (i.e. '*' where the sequence should be), instead of failing with an error", required = false, exclusiveOf = "", validation = "")
- var filter_bases_not_stored: Boolean = _
+ var filter_bases_not_stored: Boolean = config("filter_bases_not_stored", default = false)
if (config.contains("dbsnp")) knownSites :+= new File(config("dbsnp").asString)
@@ -166,7 +151,6 @@ class BaseRecalibrator(val root: Configurable) extends CommandLineGATK /* with S
override def cmdLine = super.cmdLine +
repeat("-knownSites", knownSites, formatPrefix = TaggedFile.formatCommandLineParameter, spaceSeparated = true, escape = true, format = "%s") +
required("-o", out, spaceSeparated = true, escape = true, format = "%s") +
- conditional(list, "-ls", escape = true, format = "%s") +
repeat("-cov", covariate, spaceSeparated = true, escape = true, format = "%s") +
conditional(no_standard_covs, "-noStandard", escape = true, format = "%s") +
conditional(run_without_dbsnp_potentially_ruining_quality, "-run_without_dbsnp_potentially_ruining_quality", escape = true, format = "%s") +
--
GitLab