From 8ac534fc067752fde862464f7c0c25224e4321c1 Mon Sep 17 00:00:00 2001
From: Olivier Langella <Olivier.Langella@moulon.inra.fr>
Date: Wed, 19 Apr 2017 22:29:07 +0200
Subject: [PATCH] new fasta file object and identification engine parameters
---
src/CMakeLists.txt | 1 +
.../identificationdatasource.cpp | 7 +++
.../identificationdatasource.h | 11 +++++
src/files/fastafile.cpp | 45 +++++++++++++++++++
src/files/fastafile.h | 43 ++++++++++++++++++
src/input/xtandemsaxhandler.cpp | 43 +++++++++++-------
src/utils/types.h | 7 +++
7 files changed, 140 insertions(+), 17 deletions(-)
create mode 100644 src/files/fastafile.cpp
create mode 100644 src/files/fastafile.h
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index d9c1965b4..cd161d715 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -61,6 +61,7 @@ SET(CPP_FILES
core/proteinmatch.cpp
core/proteinxtp.cpp
core/sequencedatabase.cpp
+ files/fastafile.cpp
files/xpipfile.cpp
grouping/groupingexperiment.cpp
grouping/groupinggroup.cpp
diff --git a/src/core/identification_sources/identificationdatasource.cpp b/src/core/identification_sources/identificationdatasource.cpp
index f3cc31f80..dd06ce03a 100644
--- a/src/core/identification_sources/identificationdatasource.cpp
+++ b/src/core/identification_sources/identificationdatasource.cpp
@@ -72,8 +72,15 @@ const QString& IdentificationDataSource::getIdentificationEngineVersion() const
void IdentificationDataSource::setIdentificationEngineVersion(const QString& version) {
_version = version;
}
+void IdentificationDataSource::setIdentificationEngineParam(IdentificationEngineParam param, const QString& value) {
+ _params.insert(std::pair<IdentificationEngineParam, QString>(param, value));
+}
pappso::SpectrumSp IdentificationDataSource::getSpectrumSp(unsigned int scan_number) const {
pappso::SpectrumSp spectrum_sp = SpectrumStore::getSpectrumSpFromMsRunSp(_ms_run_sp, scan_number);
return spectrum_sp;
}
+
+void IdentificationDataSource::addFastaFile (FastaFile file) {
+ _fastafile_list.push_back(file);
+}
diff --git a/src/core/identification_sources/identificationdatasource.h b/src/core/identification_sources/identificationdatasource.h
index 02d1032a5..3d4b51f2b 100644
--- a/src/core/identification_sources/identificationdatasource.h
+++ b/src/core/identification_sources/identificationdatasource.h
@@ -27,6 +27,7 @@
#include <pappsomspp/spectrum/spectrum.h>
#include <memory>
#include "../msrun.h"
+#include "../../files/fastafile.h"
class Project;
@@ -65,6 +66,14 @@ public:
/** \brief set identification engine version
*/
virtual void setIdentificationEngineVersion(const QString& version);
+
+ /** \brief set identification engine parameter value
+ */
+ virtual void setIdentificationEngineParam(IdentificationEngineParam param, const QString& value);
+
+ /** \brief add Fastafile used by the identification engine
+ */
+ void addFastaFile (FastaFile file);
protected :
QString _resource_name;
@@ -73,6 +82,8 @@ private :
//static std::map<QString, pappso::MsRunIdSp> _map_msrunidsp;
QString _version;
MsRunSp _ms_run_sp = nullptr;
+ std::map<IdentificationEngineParam, QString> _params;
+ std::vector<FastaFile> _fastafile_list;
};
#endif // IDENTIFICATIONDATASOURCE_H
diff --git a/src/files/fastafile.cpp b/src/files/fastafile.cpp
new file mode 100644
index 000000000..26d4c8358
--- /dev/null
+++ b/src/files/fastafile.cpp
@@ -0,0 +1,45 @@
+
+/*******************************************************************************
+* Copyright (c) 2017 Olivier Langella <olivier.langella@u-psud.fr>.
+*
+* This file is part of XTPcpp.
+*
+* XTPcpp is free software: you can redistribute it and/or modify
+* it under the terms of the GNU General Public License as published by
+* the Free Software Foundation, either version 3 of the License, or
+* (at your option) any later version.
+*
+* XTPcpp is distributed in the hope that it will be useful,
+* but WITHOUT ANY WARRANTY; without even the implied warranty of
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+* GNU General Public License for more details.
+*
+* You should have received a copy of the GNU General Public License
+* along with XTPcpp. If not, see <http://www.gnu.org/licenses/>.
+*
+* Contributors:
+* Olivier Langella <olivier.langella@u-psud.fr> - initial API and implementation
+******************************************************************************/
+
+#include "fastafile.h"
+
+FastaFile::FastaFile(const QString & fasta_source) : _fasta_source(fasta_source)
+{
+
+}
+
+FastaFile::FastaFile(const QUrl & fasta_source) : _fasta_source(fasta_source)
+{
+
+}
+FastaFile::FastaFile(const QFileInfo & fasta_source): _fasta_source(fasta_source.absoluteFilePath())
+{
+}
+FastaFile::FastaFile(const FastaFile & other) : _fasta_source(other._fasta_source)
+{
+
+}
+FastaFile::~FastaFile()
+{
+
+}
diff --git a/src/files/fastafile.h b/src/files/fastafile.h
new file mode 100644
index 000000000..3e3c90a39
--- /dev/null
+++ b/src/files/fastafile.h
@@ -0,0 +1,43 @@
+
+/*******************************************************************************
+* Copyright (c) 2017 Olivier Langella <olivier.langella@u-psud.fr>.
+*
+* This file is part of XTPcpp.
+*
+* XTPcpp is free software: you can redistribute it and/or modify
+* it under the terms of the GNU General Public License as published by
+* the Free Software Foundation, either version 3 of the License, or
+* (at your option) any later version.
+*
+* XTPcpp is distributed in the hope that it will be useful,
+* but WITHOUT ANY WARRANTY; without even the implied warranty of
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+* GNU General Public License for more details.
+*
+* You should have received a copy of the GNU General Public License
+* along with XTPcpp. If not, see <http://www.gnu.org/licenses/>.
+*
+* Contributors:
+* Olivier Langella <olivier.langella@u-psud.fr> - initial API and implementation
+******************************************************************************/
+
+#ifndef FASTAFILE_H
+#define FASTAFILE_H
+
+#include <QUrl>
+#include <QFileInfo>
+class FastaFile
+{
+public:
+ FastaFile(const QString & fasta_source);
+ FastaFile(const QUrl & fasta_source);
+ FastaFile(const QFileInfo & fasta_source);
+ FastaFile(const FastaFile & other);
+ ~FastaFile();
+
+private :
+ const QUrl _fasta_source;
+
+};
+
+#endif // FASTAFILE_H
diff --git a/src/input/xtandemsaxhandler.cpp b/src/input/xtandemsaxhandler.cpp
index 1ef6a3d36..23b4862b6 100644
--- a/src/input/xtandemsaxhandler.cpp
+++ b/src/input/xtandemsaxhandler.cpp
@@ -34,6 +34,7 @@
#include <cmath>
#include "../utils/peptidestore.h"
#include "../utils/proteinstore.h"
+#include "../files/fastafile.h"
XtandemSaxHandler::XtandemSaxHandler(Project * p_project, IdentificationGroup * p_identification_group,
IdentificationDataSource * p_identification_data_source):_p_project(p_project)
@@ -334,8 +335,11 @@ bool XtandemSaxHandler::endElement_note() {
else {
//<group label="input parameters" type="parameters">
+ //<note type="input" label="list path, default parameters">/gorgone/pappso/tmp/temp_condor_job8533994640337729751189420695540169/QExactive_analysis_FDR_nosemi.xml</note>
+ if (_current_note_label == "list path, default parameters") {
+ _p_identification_data_source->setIdentificationEngineParam(IdentificationEngineParam::tandem_param,_current_text);
+ }
/*
- * <note type="input" label="list path, default parameters">/gorgone/pappso/tmp/temp_condor_job8533994640337729751189420695540169/QExactive_analysis_FDR_nosemi.xml</note>
<note type="input" label="list path, taxonomy information">/gorgone/pappso/tmp/temp_condor_job8533994640337729751189420695540169/database.xml</note>
<note type="input" label="output, histogram column width">30</note>
<note type="input" label="output, histograms">yes</note>
@@ -441,23 +445,28 @@ bool XtandemSaxHandler::endElement_note() {
*/
//<group label="performance parameters" type="parameters">
+
+ //<note label="list path, sequence source #1">/gorgone/pappso/formation/TD/Database/Genome_Z_mays_5a.fasta</note>
+ //<note label="list path, sequence source #2">/gorgone/pappso/formation/TD/Database/contaminants_standarts.fasta</note>
+ if (_current_note_label.startsWith("list path, sequence source")) {
+ _p_identification_data_source->addFastaFile(FastaFile(_current_text));
+ }
+
/*
- <note label="list path, sequence source #1">/gorgone/pappso/formation/TD/Database/Genome_Z_mays_5a.fasta</note>
- <note label="list path, sequence source #2">/gorgone/pappso/formation/TD/Database/contaminants_standarts.fasta</note>
- <note label="list path, sequence source description #1">no description</note>
- <note label="list path, sequence source description #2">no description</note>
- <note label="modelling, duplicate peptide ids">6019</note>
- <note label="modelling, duplicate proteins">19735</note>
- <note label="modelling, estimated false positives">18</note>
- <note label="modelling, reversed sequence false positives">20</note>
- <note label="modelling, spectrum noise suppression ratio">0.00</note>
- <note label="modelling, total peptides used">96618641</note>
- <note label="modelling, total proteins used">273656</note>
- <note label="modelling, total spectra assigned">7464</note>
- <note label="modelling, total spectra used">12199</note>
- <note label="modelling, total unique assigned">6260</note>
- <note label="process, start time">2013:12:20:16:47:19</note>
- */
+ <note label="list path, sequence source description #1">no description</note>
+ <note label="list path, sequence source description #2">no description</note>
+ <note label="modelling, duplicate peptide ids">6019</note>
+ <note label="modelling, duplicate proteins">19735</note>
+ <note label="modelling, estimated false positives">18</note>
+ <note label="modelling, reversed sequence false positives">20</note>
+ <note label="modelling, spectrum noise suppression ratio">0.00</note>
+ <note label="modelling, total peptides used">96618641</note>
+ <note label="modelling, total proteins used">273656</note>
+ <note label="modelling, total spectra assigned">7464</note>
+ <note label="modelling, total spectra used">12199</note>
+ <note label="modelling, total unique assigned">6260</note>
+ <note label="process, start time">2013:12:20:16:47:19</note>
+ */
//<note label="process, version">X! Tandem Sledgehammer (2013.09.01.1)</note>
if (_current_note_label == "process, version") {
QRegExp rx("\\((.*)\\)");
diff --git a/src/utils/types.h b/src/utils/types.h
index 6c98714a1..aa354e8cd 100644
--- a/src/utils/types.h
+++ b/src/utils/types.h
@@ -37,6 +37,13 @@ enum class IdentificationEngine {
peptider ///< peptider
};
+/** \def IdentificationEngineParam identification engine parameters
+ *
+ */
+enum class IdentificationEngineParam {
+ tandem_param ///< X!Tandem xml parameters file
+};
+
/** \def GroupingType list of available grouping algoritms
*
*/
--
GitLab