Skip to content
Snippets Groups Projects
Commit 93a925f6 authored by Langella Olivier's avatar Langella Olivier
Browse files

write Protein ID in ODS protein sheet

parent 245bf433
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
src/output/ods/proteinsheet.cpp
+ 219
− 173
View file @ 93a925f6
......@@ -6,204 +6,250 @@
*/
/*******************************************************************************
* 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
******************************************************************************/
* 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 "proteinsheet.h"
#include "../../core/identificationgroup.h"
#include <pappsomspp/utils.h>
#include <pappsomspp/pappsoexception.h>
ProteinSheet::ProteinSheet (OdsExport * p_ods_export, CalcWriterInterface * p_writer, const Project * p_project): _p_project(p_project) {
_p_writer = p_writer;
_p_ods_export = p_ods_export;
p_writer->writeSheet("proteins");
OdsTableSettings table_settings;
table_settings.setVerticalSplit(1);
_p_writer->setCurrentOdsTableSettings(table_settings);
std::vector<IdentificationGroup *> identification_list = p_project->getIdentificationGroupList();
for (IdentificationGroup * p_ident:identification_list) {
//writeHeaders(p_ident);
writeIdentificationGroup(p_ident);
ProteinSheet::ProteinSheet(OdsExport *p_ods_export,
CalcWriterInterface *p_writer,
const Project *p_project)
: _p_project(p_project)
{
_p_writer = p_writer;
_p_ods_export = p_ods_export;
p_writer->writeSheet("proteins");
OdsTableSettings table_settings;
table_settings.setVerticalSplit(1);
_p_writer->setCurrentOdsTableSettings(table_settings);
std::vector<IdentificationGroup *> identification_list =
p_project->getIdentificationGroupList();
for(IdentificationGroup *p_ident : identification_list)
{
// writeHeaders(p_ident);
writeIdentificationGroup(p_ident);
}
}
void ProteinSheet::writeCellHeader(ProteinListColumn column) {
qDebug() << "ProteinSheet::writeCellHeader begin " << (std::int8_t) column;
_p_writer->setCellAnnotation(ProteinTableModel::getDescription(column));
_p_writer->writeCell(ProteinTableModel::getTitle(column));
qDebug() << "ProteinSheet::writeCellHeader end" ;
void
ProteinSheet::writeCellHeader(ProteinListColumn column)
{
qDebug() << "ProteinSheet::writeCellHeader begin " << (std::int8_t)column;
_p_writer->setCellAnnotation(ProteinTableModel::getDescription(column));
_p_writer->writeCell(ProteinTableModel::getTitle(column));
qDebug() << "ProteinSheet::writeCellHeader end";
}
void ProteinSheet::writeHeaders(IdentificationGroup * p_ident) {
// Group ID Sub-group ID Protein ID Description log(E value)
// Coverage MW Spectra Specific uniques Uniques PAI Redundancy
const std::vector<MsRunSp> msrun_list = p_ident->getMsRunSpList();
if (msrun_list.size() == 1) {
_p_writer->writeCell("sample");
_p_writer->writeLine();
_p_writer->writeCell(msrun_list[0].get()->getSampleName());
_p_writer->writeLine();
void
ProteinSheet::writeHeaders(IdentificationGroup *p_ident)
{
// Group ID Sub-group ID Protein ID Description log(E value)
// Coverage MW Spectra Specific uniques Uniques PAI Redundancy
const std::vector<MsRunSp> msrun_list = p_ident->getMsRunSpList();
if(msrun_list.size() == 1)
{
_p_writer->writeCell("sample");
_p_writer->writeLine();
_p_writer->writeCell(msrun_list[0].get()->getSampleName());
_p_writer->writeLine();
}
_p_writer->writeLine();
_p_writer->setCellAnnotation("group number");
_p_writer->writeCell("Group ID");
_p_writer->setCellAnnotation("subgroup number");
_p_writer->writeCell("Sub-group ID");
writeCellHeader(ProteinListColumn::protein_grouping_id);
//_p_writer->setCellAnnotation("unique protein identifier within this grouping experiment");
//_p_writer->writeCell("Protein ID");
writeCellHeader(ProteinListColumn::accession);
writeCellHeader(ProteinListColumn::description);
writeCellHeader(ProteinListColumn::log_evalue);
writeCellHeader(ProteinListColumn::coverage);
_p_writer->setCellAnnotation("computed molecular weight for this protein (sum of amino acid masses)");
_p_writer->writeCell("MW");
_p_writer->setCellAnnotation("number of amino acids");
_p_writer->writeCell("size");
writeCellHeader(ProteinListColumn::spectrum);
//_p_writer->setCellAnnotation("number of scans (spectra) attributed to this protein");
// _p_writer->writeCell("Spectra");
writeCellHeader(ProteinListColumn::specific_spectrum);
writeCellHeader(ProteinListColumn::sequence);
writeCellHeader(ProteinListColumn::specific_sequence);
_p_writer->setCellAnnotation("number of unique peptidesequenceLI+modification+charge attributed to this protein (used to compute the PAI)");
_p_writer->writeCell("Uniques peptide-mod-charge");
_p_writer->setCellAnnotation("theoretical number of tryptic peptides used in the PAI computation");
_p_writer->writeCell("Theoretical number of tryptic peptides");
writeCellHeader(ProteinListColumn::pai);
writeCellHeader(ProteinListColumn::empai);
//_p_writer->writeCell("spectral count AI");
_p_writer->setCellAnnotation("number of proteins in the same subgroup");
_p_writer->writeCell("Sub-group proteins");
_p_writer->setCellAnnotation("number of MS samples in which this protein was observed");
_p_writer->writeCell("number of MS samples");
_p_writer->writeLine();
_p_writer->setCellAnnotation("group number");
_p_writer->writeCell("Group ID");
_p_writer->setCellAnnotation("subgroup number");
_p_writer->writeCell("Sub-group ID");
_p_writer->setCellAnnotation(ProteinTableModel::getDescription(ProteinListColumn::protein_grouping_id));
_p_writer->writeCell("Protein ID");
//writeCellHeader(ProteinListColumn::protein_grouping_id);
//_p_writer->setCellAnnotation("unique protein identifier within this grouping
//experiment"); _p_writer->writeCell("Protein ID");
writeCellHeader(ProteinListColumn::accession);
writeCellHeader(ProteinListColumn::description);
writeCellHeader(ProteinListColumn::log_evalue);
writeCellHeader(ProteinListColumn::coverage);
_p_writer->setCellAnnotation(
"computed molecular weight for this protein (sum of amino acid masses)");
_p_writer->writeCell("MW");
_p_writer->setCellAnnotation("number of amino acids");
_p_writer->writeCell("size");
writeCellHeader(ProteinListColumn::spectrum);
//_p_writer->setCellAnnotation("number of scans (spectra) attributed to this
//protein");
// _p_writer->writeCell("Spectra");
writeCellHeader(ProteinListColumn::specific_spectrum);
writeCellHeader(ProteinListColumn::sequence);
writeCellHeader(ProteinListColumn::specific_sequence);
_p_writer->setCellAnnotation(
"number of unique peptidesequenceLI+modification+charge attributed to this "
"protein (used to compute the PAI)");
_p_writer->writeCell("Uniques peptide-mod-charge");
_p_writer->setCellAnnotation(
"theoretical number of tryptic peptides used in the PAI computation");
_p_writer->writeCell("Theoretical number of tryptic peptides");
writeCellHeader(ProteinListColumn::pai);
writeCellHeader(ProteinListColumn::empai);
//_p_writer->writeCell("spectral count AI");
_p_writer->setCellAnnotation("number of proteins in the same subgroup");
_p_writer->writeCell("Sub-group proteins");
_p_writer->setCellAnnotation(
"number of MS samples in which this protein was observed");
_p_writer->writeCell("number of MS samples");
}
void ProteinSheet::writeIdentificationGroup(IdentificationGroup * p_ident) {
void
ProteinSheet::writeIdentificationGroup(IdentificationGroup *p_ident)
{
writeHeaders(p_ident);
for (const std::pair<unsigned int, GroupingGroupSp> & group_pair : p_ident->getGroupStore().getGroupMap()) {
writeHeaders(p_ident);
for(const std::pair<unsigned int, GroupingGroupSp> &group_pair :
p_ident->getGroupStore().getGroupMap())
{
std::vector<const ProteinMatch *> protein_match_list = group_pair.second.get()->getProteinMatchList();
std::vector<const ProteinMatch *> protein_match_list =
group_pair.second.get()->getProteinMatchList();
std::sort(protein_match_list.begin(), protein_match_list.end(),
[](const ProteinMatch * a, const ProteinMatch * b)
{
return a->getGrpProteinSp().get()->getSubGroupNumber() < b->getGrpProteinSp().get()->getSubGroupNumber();
});
std::sort(protein_match_list.begin(), protein_match_list.end(),
[](const ProteinMatch *a, const ProteinMatch *b) {
return a->getGrpProteinSp().get()->getSubGroupNumber() <
b->getGrpProteinSp().get()->getSubGroupNumber();
});
for (auto & protein_match:protein_match_list) {
writeOneProtein(group_pair.second.get(), protein_match);
for(auto &protein_match : protein_match_list)
{
writeOneProtein(group_pair.second.get(), protein_match);
}
}
_p_writer->writeLine();
_p_writer->writeLine();
}
void ProteinSheet::writeOneProtein(const GroupingGroup * p_group, const ProteinMatch * p_protein_match) {
try {
_p_writer->writeLine();
qDebug() << "ProteinSheet::writeOneProtein begin" ;
ValidationState validation_state = ValidationState::validAndChecked;
pappso::GrpProtein * p_grp_protein = p_protein_match->getGrpProteinSp().get();
ProteinXtp * p_protein = p_protein_match->getProteinXtpSp().get();
unsigned int group_number = p_grp_protein->getGroupNumber();
unsigned int subgroup_number = p_grp_protein->getSubGroupNumber();
_p_ods_export->setEvenOrOddStyle(group_number, _p_writer);
_p_writer->writeCell(pappso::Utils::getLexicalOrderedString(group_number));
_p_ods_export->setEvenOrOddStyle(subgroup_number, _p_writer);
_p_writer->writeCell(QString("%1.%2").arg(pappso::Utils::getLexicalOrderedString(group_number)).arg(pappso::Utils::getLexicalOrderedString(subgroup_number)));
_p_writer->clearTableCellStyleRef();
_p_writer->writeCell(p_grp_protein->getGroupingId());
const std::list<DbXref> & dbxref_list = p_protein->getDbxrefList();
if (dbxref_list.size() == 0) {
_p_writer->writeCell(p_protein->getAccession());
void
ProteinSheet::writeOneProtein(const GroupingGroup *p_group,
const ProteinMatch *p_protein_match)
{
try
{
_p_writer->writeLine();
qDebug() << "ProteinSheet::writeOneProtein begin";
ValidationState validation_state = ValidationState::validAndChecked;
pappso::GrpProtein *p_grp_protein =
p_protein_match->getGrpProteinSp().get();
ProteinXtp *p_protein = p_protein_match->getProteinXtpSp().get();
unsigned int group_number = p_grp_protein->getGroupNumber();
unsigned int subgroup_number = p_grp_protein->getSubGroupNumber();
_p_ods_export->setEvenOrOddStyle(group_number, _p_writer);
_p_writer->writeCell(
pappso::Utils::getLexicalOrderedString(group_number));
_p_ods_export->setEvenOrOddStyle(subgroup_number, _p_writer);
_p_writer->writeCell(
QString("%1.%2")
.arg(pappso::Utils::getLexicalOrderedString(group_number))
.arg(pappso::Utils::getLexicalOrderedString(subgroup_number)));
_p_writer->clearTableCellStyleRef();
_p_writer->writeCell(p_grp_protein->getGroupingId());
const std::list<DbXref> &dbxref_list = p_protein->getDbxrefList();
if(dbxref_list.size() == 0)
{
_p_writer->writeCell(p_protein->getAccession());
}
else {
_p_writer->writeCell(dbxref_list.front().getUrl(), p_protein->getAccession());
else
{
_p_writer->writeCell(dbxref_list.front().getUrl(),
p_protein->getAccession());
}
_p_writer->writeCell(p_protein->getDescription());
_p_writer->writeCell(p_protein_match->getLogEvalue());
// _p_writer->writeCell("Coverage");
_p_writer->writeCellPercentage(p_protein_match->getCoverage());
// _p_writer->writeCell("MW");
_p_writer->writeCell(p_protein->getMass());
_p_writer->writeCell(p_protein->size());
// _p_writer->writeCell("Spectra");
_p_writer->writeCell(p_protein_match->countSampleScan(validation_state));
// _p_writer->writeCell("Specific");
_p_writer->writeCell((unsigned int) p_group->countSpecificSampleScan(p_protein_match, validation_state));
// _p_writer->writeCell("Uniques");
_p_writer->writeCell((unsigned int) p_protein_match->countSequenceLi(validation_state));
// _p_writer->writeCell("Specific uniques");
_p_writer->writeCell((unsigned int) p_group->countSpecificSequenceLi(p_protein_match, validation_state));
_p_writer->writeCell(p_protein_match->countPeptideMassCharge(validation_state));
qDebug() << "ProteinSheet::writeOneProtein tryptic" ;
_p_writer->writeCell(p_protein->countTrypticPeptidesForPAI());
// _p_writer->writeCell("PAI");
// _p_writer->writeCell(new Double((float) hashProt
// .getTotalHashSampleScan() / prot.get_PAI_count()));
_p_writer->writeCell(p_protein_match->getPAI());
// _p_writer->writeCell(match.getPAI());//this is the same
/*
* ancien PAI _p_writer->writeCell(new Double((float)
* top.get_spectra_numbers_to(sample) /
* top.get_protein_match().get_PAI_count()));
*/
_p_writer->writeCell(p_protein_match->getEmPAI());
//_p_writer->writeCell(p_protein_match->getSpectralCountAI());
// _p_writer->writeCell("Redundancy");
_p_writer->writeCell(p_group->countProteinInSubgroup(p_grp_protein->getSubGroupNumber()));
//number of MS sample
_p_writer->writeCell(p_protein_match->countDistinctMsSamples(validation_state));
qDebug() << "ProteinSheet::writeOneProtein end" ;
_p_writer->writeCell(p_protein->getDescription());
_p_writer->writeCell(p_protein_match->getLogEvalue());
// _p_writer->writeCell("Coverage");
_p_writer->writeCellPercentage(p_protein_match->getCoverage());
// _p_writer->writeCell("MW");
_p_writer->writeCell(p_protein->getMass());
_p_writer->writeCell(p_protein->size());
// _p_writer->writeCell("Spectra");
_p_writer->writeCell(p_protein_match->countSampleScan(validation_state));
// _p_writer->writeCell("Specific");
_p_writer->writeCell((unsigned int)p_group->countSpecificSampleScan(
p_protein_match, validation_state));
// _p_writer->writeCell("Uniques");
_p_writer->writeCell(
(unsigned int)p_protein_match->countSequenceLi(validation_state));
// _p_writer->writeCell("Specific uniques");
_p_writer->writeCell((unsigned int)p_group->countSpecificSequenceLi(
p_protein_match, validation_state));
_p_writer->writeCell(
p_protein_match->countPeptideMassCharge(validation_state));
qDebug() << "ProteinSheet::writeOneProtein tryptic";
_p_writer->writeCell(p_protein->countTrypticPeptidesForPAI());
// _p_writer->writeCell("PAI");
// _p_writer->writeCell(new Double((float) hashProt
// .getTotalHashSampleScan() / prot.get_PAI_count()));
_p_writer->writeCell(p_protein_match->getPAI());
// _p_writer->writeCell(match.getPAI());//this is the same
/*
* ancien PAI _p_writer->writeCell(new Double((float)
* top.get_spectra_numbers_to(sample) /
* top.get_protein_match().get_PAI_count()));
*/
_p_writer->writeCell(p_protein_match->getEmPAI());
//_p_writer->writeCell(p_protein_match->getSpectralCountAI());
// _p_writer->writeCell("Redundancy");
_p_writer->writeCell(
p_group->countProteinInSubgroup(p_grp_protein->getSubGroupNumber()));
// number of MS sample
_p_writer->writeCell(
p_protein_match->countDistinctMsSamples(validation_state));
qDebug() << "ProteinSheet::writeOneProtein end";
}
catch (pappso::PappsoException error) {
throw pappso::PappsoException(QObject::tr("Error writing protein %1 :\n%2").arg(p_protein_match->getProteinXtpSp().get()->getAccession()).arg(error.qwhat()));
catch(pappso::PappsoException error)
{
throw pappso::PappsoException(
QObject::tr("Error writing protein %1 :\n%2")
.arg(p_protein_match->getProteinXtpSp().get()->getAccession())
.arg(error.qwhat()));
}
}
src/output/ods/proteinsheet.h
+ 38
− 33
View file @ 93a925f6
......@@ -6,26 +6,27 @@
*/
/*******************************************************************************
* 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
******************************************************************************/
* 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 PROTEINSHEET_H
#define PROTEINSHEET_H
......@@ -38,19 +39,23 @@
class ProteinSheet
{
public :
ProteinSheet (OdsExport * p_ods_export, CalcWriterInterface * p_writer, const Project * p_project);
private :
void writeIdentificationGroup(IdentificationGroup * p_ident);
void writeHeaders(IdentificationGroup * p_ident);
void writeOneProtein(const GroupingGroup * p_group, const ProteinMatch * p_protein_match);
protected:
void writeCellHeader(ProteinListColumn column);
private :
OdsExport * _p_ods_export;
const Project * _p_project;
CalcWriterInterface * _p_writer;
public:
ProteinSheet(OdsExport *p_ods_export, CalcWriterInterface *p_writer,
const Project *p_project);
private:
void writeIdentificationGroup(IdentificationGroup *p_ident);
void writeHeaders(IdentificationGroup *p_ident);
void writeOneProtein(const GroupingGroup *p_group,
const ProteinMatch *p_protein_match);
protected:
void writeCellHeader(ProteinListColumn column);
private:
OdsExport *_p_ods_export;
const Project *_p_project;
CalcWriterInterface *_p_writer;
};
......
src/utils/msrunstatisticshandler.h
+ 14
− 13
View file @ 93a925f6
......@@ -32,17 +32,18 @@
class MsRunStatisticsHandler : public pappso::SpectrumCollectionHandlerInterface
{
public:
virtual void setQualifiedSpectrum (const pappso::QualifiedSpectrum & qspectrum) override;
virtual bool needPeakList () const override ;
unsigned long getMsLevelCount(unsigned int ms_level) const;
pappso::pappso_double getMsLevelTic(unsigned int ms_level) const;
unsigned long getTotalCount() const;
private:
std::vector<unsigned long> _count_ms_level_spectrum;
std::vector<pappso::pappso_double> _tic_ms_level_spectrum;
public:
virtual void
setQualifiedSpectrum(const pappso::QualifiedSpectrum &qspectrum) override;
virtual bool needPeakList() const override;
unsigned long getMsLevelCount(unsigned int ms_level) const;
pappso::pappso_double getMsLevelTic(unsigned int ms_level) const;
unsigned long getTotalCount() const;
private:
std::vector<unsigned long> _count_ms_level_spectrum;
std::vector<pappso::pappso_double> _tic_ms_level_spectrum;
};
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment