double SimplePairFinder::similarity_(ConsensusFeature const & left, ConsensusFeature const & right) const
{
double right_intensity(right.getIntensity());
if (right_intensity == 0)
return 0;
double intensity_ratio = left.getIntensity() / right_intensity;
if (intensity_ratio > 1.)
intensity_ratio = 1. / intensity_ratio;
// if the right map is the transformed map, take the transformed right position
DPosition<2> position_difference = left.getPosition() - right.getPosition();
for (UInt dimension = 0; dimension < 2; ++dimension)
{
// the formula is explained in class doc
if (position_difference[dimension] < 0)
{
position_difference[dimension] = -position_difference[dimension];
}
position_difference[dimension] *= diff_intercept_[dimension];
position_difference[dimension] += 1.0;
position_difference[dimension] = pow(position_difference[dimension], diff_exponent_[dimension]);
}
return intensity_ratio / position_difference[Peak2D::RT] / position_difference[Peak2D::MZ];
}
bool StablePairFinder::compatibleIDs_(const ConsensusFeature& feat1, const ConsensusFeature& feat2) const
{
// a feature without identifications always matches:
if (feat1.getPeptideIdentifications().empty() || feat2.getPeptideIdentifications().empty())
return true;
const vector<PeptideIdentification>& pep1 = feat1.getPeptideIdentifications();
const vector<PeptideIdentification>& pep2 = feat2.getPeptideIdentifications();
set<String> best1, best2;
for (vector<PeptideIdentification>::const_iterator pep_it = pep1.begin(); pep_it != pep1.end(); ++pep_it)
{
if (pep_it->getHits().empty())
continue; // shouldn't be the case
best1.insert(getBestHitSequence_(*pep_it).toString());
}
for (vector<PeptideIdentification>::const_iterator pep_it = pep2.begin(); pep_it != pep2.end(); ++pep_it)
{
if (pep_it->getHits().empty())
continue; // shouldn't be the case
best2.insert(getBestHitSequence_(*pep_it).toString());
}
return best1 == best2;
}
void FeatureGroupingAlgorithm::transferSubelements(const vector<ConsensusMap>& maps, ConsensusMap& out) const
{
// accumulate file descriptions from the input maps:
// cout << "Updating file descriptions..." << endl;
out.getFileDescriptions().clear();
// mapping: (map index, original id) -> new id
map<pair<Size, UInt64>, Size> mapid_table;
for (Size i = 0; i < maps.size(); ++i)
{
const ConsensusMap& consensus = maps[i];
for (ConsensusMap::FileDescriptions::const_iterator desc_it = consensus.getFileDescriptions().begin(); desc_it != consensus.getFileDescriptions().end(); ++desc_it)
{
Size counter = mapid_table.size();
mapid_table[make_pair(i, desc_it->first)] = counter;
out.getFileDescriptions()[counter] = desc_it->second;
}
}
// look-up table: input map -> unique ID -> consensus feature
// cout << "Creating look-up table..." << endl;
vector<map<UInt64, ConsensusMap::ConstIterator> > feat_lookup(maps.size());
for (Size i = 0; i < maps.size(); ++i)
{
const ConsensusMap& consensus = maps[i];
for (ConsensusMap::ConstIterator feat_it = consensus.begin();
feat_it != consensus.end(); ++feat_it)
{
// do NOT use "id_lookup[i][feat_it->getUniqueId()] = feat_it;" here as
// you will get "attempt to copy-construct an iterator from a singular
// iterator" in STL debug mode:
feat_lookup[i].insert(make_pair(feat_it->getUniqueId(), feat_it));
}
}
// adjust the consensus features:
// cout << "Adjusting consensus features..." << endl;
for (ConsensusMap::iterator cons_it = out.begin(); cons_it != out.end(); ++cons_it)
{
ConsensusFeature adjusted = ConsensusFeature(
static_cast<BaseFeature>(*cons_it)); // remove sub-features
for (ConsensusFeature::HandleSetType::const_iterator sub_it = cons_it->getFeatures().begin(); sub_it != cons_it->getFeatures().end(); ++sub_it)
{
UInt64 id = sub_it->getUniqueId();
Size map_index = sub_it->getMapIndex();
ConsensusMap::ConstIterator origin = feat_lookup[map_index][id];
for (ConsensusFeature::HandleSetType::const_iterator handle_it = origin->getFeatures().begin(); handle_it != origin->getFeatures().end(); ++handle_it)
{
FeatureHandle handle = *handle_it;
Size new_id = mapid_table[make_pair(map_index, handle.getMapIndex())];
handle.setMapIndex(new_id);
adjusted.insert(handle);
}
}
*cons_it = adjusted;
}
}
bool IsobaricChannelExtractor::hasLowIntensityReporter_(const ConsensusFeature& cf) const
{
for (ConsensusFeature::const_iterator cf_it = cf.begin();
cf_it != cf.end();
++cf_it)
{
if (cf_it->getIntensity() == 0.0)
{
return true;
}
}
return false;
}
void MapAlignmentTransformer::applyToConsensusFeature_(
ConsensusFeature& feature, const TransformationDescription& trafo,
bool store_original_rt)
{
applyToBaseFeature_(feature, trafo, store_original_rt);
// apply to grouped features (feature handles):
for (ConsensusFeature::HandleSetType::const_iterator it =
feature.getFeatures().begin(); it != feature.getFeatures().end();
++it)
{
double rt = it->getRT();
it->asMutable().setRT(trafo.apply(rt));
}
}
void MapAlignmentTransformer::applyToConsensusFeature_(ConsensusFeature & feature,
const TransformationDescription & trafo)
{
typedef ConsensusFeature::HandleSetType::const_iterator TConstHandleSetIterator;
applyToBaseFeature_(feature, trafo);
// apply to grouped features (feature handles):
for (TConstHandleSetIterator it = feature.getFeatures().begin();
it != feature.getFeatures().end();
++it)
{
DoubleReal rt = it->getRT();
it->asMutable().setRT(trafo.apply(rt));
}
}
void EDTAFile::store(const String& filename, const ConsensusMap& map) const
{
TextFile tf;
// search for maximum number of sub-features (since this determines the number of columns)
Size max_sub(0);
for (Size i = 0; i < map.size(); ++i)
{
max_sub = std::max(max_sub, map[i].getFeatures().size());
}
// write header
String header("RT\tm/z\tintensity\tcharge");
for (Size i = 1; i <= max_sub; ++i)
{
header += "\tRT" + String(i) + "\tm/z" + String(i) + "\tintensity" + String(i) + "\tcharge" + String(i);
}
tf.addLine(header);
for (Size i = 0; i < map.size(); ++i)
{
ConsensusFeature f = map[i];
// consensus
String entry = String(f.getRT()) + "\t" + f.getMZ() + "\t" + f.getIntensity() + "\t" + f.getCharge();
// sub-features
ConsensusFeature::HandleSetType handle = f.getFeatures();
for (ConsensusFeature::HandleSetType::const_iterator it = handle.begin(); it != handle.end(); ++it)
{
entry += String("\t") + it->getRT() + "\t" + it->getMZ() + "\t" + it->getIntensity() + "\t" + it->getCharge();
}
// missing sub-features
for (Size j = handle.size(); j < max_sub; ++j)
{
entry += "\tNA\tNA\tNA\tNA";
}
tf.addLine(entry);
}
tf.store(filename);
}
void MetaDataBrowser::add(ConsensusFeature & feature)
{
//peptide ids
for (std::vector<PeptideIdentification>::iterator it = feature.getPeptideIdentifications().begin(); it != feature.getPeptideIdentifications().end(); ++it)
{
add(*it);
}
add(static_cast<MetaInfoInterface &>(feature));
treeview_->expandItem(treeview_->findItems(QString::number(0), Qt::MatchExactly, 1).first());
}
ConsensusFeature getCFWithIntensites(double v[])
{
ConsensusFeature cf;
BaseFeature bf0, bf1, bf2, bf3;
bf0.setIntensity(v[0]);
bf1.setIntensity(v[1]);
bf2.setIntensity(v[2]);
bf3.setIntensity(v[3]);
cf.insert(0, bf0);cf.insert(1, bf1);cf.insert(2, bf2);cf.insert(3, bf3);
cf.setIntensity(v[0]+v[1]+v[2]+v[3]);
return cf;
}
void IsobaricChannelExtractor::extractChannels(const MSExperiment<Peak1D>& ms_exp_data, ConsensusMap& consensus_map)
{
if (ms_exp_data.empty())
{
LOG_WARN << "The given file does not contain any conventional peak data, but might"
" contain chromatograms. This tool currently cannot handle them, sorry.\n";
throw Exception::MissingInformation(__FILE__, __LINE__, __PRETTY_FUNCTION__, "Experiment has no scans!");
}
// clear the output map
consensus_map.clear(false);
consensus_map.setExperimentType("labeled_MS2");
// create predicate for spectrum checking
LOG_INFO << "Selecting scans with activation mode: " << (selected_activation_ == "" ? "any" : selected_activation_) << "\n";
HasActivationMethod<MSExperiment<Peak1D>::SpectrumType> activation_predicate(StringList::create(selected_activation_));
// now we have picked data
// --> assign peaks to channels
UInt64 element_index(0);
// remember the current precusor spectrum
MSExperiment<Peak1D>::ConstIterator prec_spec = ms_exp_data.end();
for (MSExperiment<Peak1D>::ConstIterator it = ms_exp_data.begin(); it != ms_exp_data.end(); ++it)
{
// remember the last MS1 spectra as we assume it to be the precursor spectrum
if (it->getMSLevel() == 1) prec_spec = it;
if (selected_activation_ == "" || activation_predicate(*it))
{
// check if precursor is available
if (it->getPrecursors().empty())
{
throw Exception::MissingInformation(__FILE__, __LINE__, __PRETTY_FUNCTION__, String("No precursor information given for scan native ID ") + it->getNativeID() + " with RT " + String(it->getRT()));
}
// check precursor constraints
if (!isValidPrecursor_(it->getPrecursors()[0]))
{
LOG_DEBUG << "Skip spectrum " << it->getNativeID() << ": Precursor doesn't fulfill all constraints." << std::endl;
continue;
}
// check precursor purity if we have a valid precursor ..
if (prec_spec != ms_exp_data.end())
{
const DoubleReal purity = computePrecursorPurity_(it, prec_spec);
if (purity < min_precursor_purity_)
{
LOG_DEBUG << "Skip spectrum " << it->getNativeID() << ": Precursor purity is below the threshold. [purity = " << purity << "]" << std::endl;
continue;
}
}
else
{
LOG_INFO << "No precursor available for spectrum: " << it->getNativeID() << std::endl;
}
if (!(prec_spec == ms_exp_data.end()) && computePrecursorPurity_(it, prec_spec) < min_precursor_purity_)
{
LOG_DEBUG << "Skip spectrum " << it->getNativeID() << ": Precursor purity is below the threshold." << std::endl;
continue;
}
// store RT&MZ of parent ion as centroid of ConsensusFeature
ConsensusFeature cf;
cf.setUniqueId();
cf.setRT(it->getRT());
cf.setMZ(it->getPrecursors()[0].getMZ());
Peak2D channel_value;
channel_value.setRT(it->getRT());
// for each each channel
UInt64 map_index = 0;
Peak2D::IntensityType overall_intensity = 0;
for (IsobaricQuantitationMethod::IsobaricChannelList::const_iterator cl_it = quant_method_->getChannelInformation().begin();
cl_it != quant_method_->getChannelInformation().end();
++cl_it)
{
// set mz-position of channel
channel_value.setMZ(cl_it->center);
// reset intensity
channel_value.setIntensity(0);
// as every evaluation requires time, we cache the MZEnd iterator
const MSExperiment<Peak1D>::SpectrumType::ConstIterator mz_end = it->MZEnd(cl_it->center + reporter_mass_shift_);
// add up all signals
for (MSExperiment<Peak1D>::SpectrumType::ConstIterator mz_it = it->MZBegin(cl_it->center - reporter_mass_shift_);
mz_it != mz_end;
++mz_it)
{
channel_value.setIntensity(channel_value.getIntensity() + mz_it->getIntensity());
}
// discard contribution of this channel as it is below the required intensity threshold
if (channel_value.getIntensity() < min_reporter_intensity_)
{
channel_value.setIntensity(0);
}
overall_intensity += channel_value.getIntensity();
// add channel to ConsensusFeature
cf.insert(map_index++, channel_value, element_index);
} // ! channel_iterator
// check if we keep this feature or if it contains low-intensity quantifications
if (remove_low_intensity_quantifications_ && hasLowIntensityReporter_(cf))
{
continue;
}
// check featureHandles are not empty
if (overall_intensity == 0)
{
cf.setMetaValue("all_empty", String("true"));
}
cf.setIntensity(overall_intensity);
consensus_map.push_back(cf);
// the tandem-scan in the order they appear in the experiment
++element_index;
}
} // ! Experiment iterator
/// add meta information to the map
registerChannelsInOutputMap_(consensus_map);
}
Feature feat1; Feature feat2; Feature feat3; PositionType pos1(0,0); PositionType pos2(200,300); PositionType pos3(400,500); feat1.setPosition(pos1); feat1.setIntensity(100.0f); feat1.setUniqueId(0); feat2.setPosition(pos2); feat2.setIntensity(300.0f); feat2.setUniqueId(1); feat3.setPosition(pos3); feat3.setIntensity(400.0f); feat3.setUniqueId(2); ConsensusFeature cons1(0,feat1); ConsensusFeature cons2(0,feat2); ConsensusFeature cons3(0,feat3); input[0].push_back(cons1); input[0].push_back(cons2); input[0].push_back(cons3); Feature feat4; Feature feat5; Feature feat6; PositionType pos4(4,0.04); PositionType pos5(204,300.04); PositionType pos6(404,500.04); feat4.setPosition(pos4); feat4.setIntensity(100.0f); feat4.setUniqueId(0);
void SimplePairFinder::run(const std::vector<ConsensusMap> & input_maps, ConsensusMap & result_map)
{
if (input_maps.size() != 2)
throw Exception::IllegalArgument(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, "exactly two input maps required");
checkIds_(input_maps);
// progress dots
Int progress_dots = 0;
if (this->param_.exists("debug::progress_dots"))
{
progress_dots = (Int) this->param_.getValue("debug:progress_dots");
}
Int number_of_considered_element_pairs = 0;
// For each element in map 0, find its best friend in map 1
std::vector<UInt> best_companion_index_0(input_maps[0].size(), UInt(-1));
std::vector<double> best_companion_quality_0(input_maps[0].size(), 0);
for (UInt fi0 = 0; fi0 < input_maps[0].size(); ++fi0)
{
double best_quality = -std::numeric_limits<double>::max();
for (UInt fi1 = 0; fi1 < input_maps[1].size(); ++fi1)
{
double quality = similarity_(input_maps[0][fi0], input_maps[1][fi1]);
if (quality > best_quality)
{
best_quality = quality;
best_companion_index_0[fi0] = fi1;
}
++number_of_considered_element_pairs;
if (progress_dots && !(number_of_considered_element_pairs % progress_dots))
{
std::cout << '-' << std::flush;
}
}
best_companion_quality_0[fi0] = best_quality;
}
// For each element in map 1, find its best friend in map 0
std::vector<UInt> best_companion_index_1(input_maps[1].size(), UInt(-1));
std::vector<double> best_companion_quality_1(input_maps[1].size(), 0);
for (UInt fi1 = 0; fi1 < input_maps[1].size(); ++fi1)
{
double best_quality = -std::numeric_limits<double>::max();
for (UInt fi0 = 0; fi0 < input_maps[0].size(); ++fi0)
{
double quality = similarity_(input_maps[0][fi0], input_maps[1][fi1]);
if (quality > best_quality)
{
best_quality = quality;
best_companion_index_1[fi1] = fi0;
}
++number_of_considered_element_pairs;
if (progress_dots && !(number_of_considered_element_pairs % progress_dots))
{
std::cout << '+' << std::flush;
}
}
best_companion_quality_1[fi1] = best_quality;
}
// And if both like each other, they become a pair.
// element_pairs_->clear();
for (UInt fi0 = 0; fi0 < input_maps[0].size(); ++fi0)
{
// fi0 likes someone ...
if (best_companion_quality_0[fi0] > pair_min_quality_)
{
// ... who likes him too ...
UInt best_companion_of_fi0 = best_companion_index_0[fi0];
if (best_companion_index_1[best_companion_of_fi0] == fi0 &&
best_companion_quality_1[best_companion_of_fi0] > pair_min_quality_
)
{
ConsensusFeature f;
f.insert(input_maps[0][fi0]);
f.insert(input_maps[1][best_companion_of_fi0]);
f.computeConsensus();
f.setQuality(best_companion_quality_0[fi0] + best_companion_quality_1[best_companion_of_fi0]);
result_map.push_back(f);
}
}
}
return;
}
TEST_EQUAL(map[0].getPeptideIdentifications()[1].getHits()[0].getSequence(), "C")
TEST_EQUAL(map[0].getPeptideIdentifications()[1].getHits()[1].getSequence(), "D")
TEST_EQUAL(map[1].getPeptideIdentifications().size(), 1)
TEST_EQUAL(map[1].getPeptideIdentifications()[0].getHits().size(), 1)
TEST_EQUAL(map[1].getPeptideIdentifications()[0].getHits()[0].getSequence(), "E")
//unassigned peptide identifications
TEST_EQUAL(map.getUnassignedPeptideIdentifications().size(), 2)
TEST_EQUAL(map.getUnassignedPeptideIdentifications()[0].getHits().size(), 1)
TEST_EQUAL(map.getUnassignedPeptideIdentifications()[0].getHits()[0].getSequence(), "F")
TEST_EQUAL(map.getUnassignedPeptideIdentifications()[1].getHits().size(), 2)
TEST_EQUAL(map.getUnassignedPeptideIdentifications()[1].getHits()[0].getSequence(), "G")
TEST_EQUAL(map.getUnassignedPeptideIdentifications()[1].getHits()[1].getSequence(), "H")
//features
TEST_EQUAL(map.size(), 6)
ConsensusFeature cons_feature = map[0];
TEST_REAL_SIMILAR(cons_feature.getRT(), 1273.27)
TEST_REAL_SIMILAR(cons_feature.getMZ(), 904.47)
TEST_REAL_SIMILAR(cons_feature.getIntensity(), 3.12539e+07)
TEST_REAL_SIMILAR(cons_feature.getPositionRange().minPosition()[0], 1273.27)
TEST_REAL_SIMILAR(cons_feature.getPositionRange().maxPosition()[0], 1273.27)
TEST_REAL_SIMILAR(cons_feature.getPositionRange().minPosition()[1], 904.47)
TEST_REAL_SIMILAR(cons_feature.getPositionRange().maxPosition()[1], 904.47)
TEST_REAL_SIMILAR(cons_feature.getIntensityRange().minPosition()[0], 3.12539e+07)
TEST_REAL_SIMILAR(cons_feature.getIntensityRange().maxPosition()[0], 3.12539e+07)
TEST_REAL_SIMILAR(cons_feature.getQuality(), 1.1)
TEST_EQUAL(cons_feature.getMetaValue("peptide_id") == DataValue("RefSeq:NC_1234"), true)
ConsensusFeature::HandleSetType::const_iterator it = cons_feature.begin();
TEST_REAL_SIMILAR(it->getIntensity(), 3.12539e+07)
cons_feature = map[5];
void EDTAFile::load(const String& filename, ConsensusMap& consensus_map)
{
// load input
TextFile input(filename);
TextFile::ConstIterator input_it = input.begin();
// reset map
consensus_map = ConsensusMap();
consensus_map.setUniqueId();
char separator = ' ';
if (input_it->hasSubstring("\t"))
separator = '\t';
else if (input_it->hasSubstring(" "))
separator = ' ';
else if (input_it->hasSubstring(","))
separator = ',';
// parsing header line
std::vector<String> headers;
input_it->split(separator, headers);
int offset = 0;
for (Size i = 0; i < headers.size(); ++i)
{
headers[i].trim();
}
String header_trimmed = *input.begin();
header_trimmed.trim();
enum
{
TYPE_UNDEFINED,
TYPE_OLD_NOCHARGE,
TYPE_OLD_CHARGE,
TYPE_CONSENSUS
}
input_type = TYPE_UNDEFINED;
Size input_features = 1;
double rt = 0.0;
double mz = 0.0;
double it = 0.0;
Int ch = 0;
if (headers.size() <= 2)
{
throw Exception::ParseError(__FILE__, __LINE__, __PRETTY_FUNCTION__, "", String("Failed parsing in line 1: not enough columns! Expected at least 3 columns!\nOffending line: '") + header_trimmed + "' (line 1)\n");
}
else if (headers.size() == 3)
input_type = TYPE_OLD_NOCHARGE;
else if (headers.size() == 4)
input_type = TYPE_OLD_CHARGE;
// see if we have a header
try
{
// try to convert... if not: thats a header
rt = headers[0].toDouble();
mz = headers[1].toDouble();
it = headers[2].toDouble();
}
catch (Exception::BaseException&)
{
offset = 1;
++input_it;
LOG_INFO << "Detected a header line.\n";
}
if (headers.size() >= 5)
{
if (String(headers[4].trim()).toUpper() == "RT1")
input_type = TYPE_CONSENSUS;
else
input_type = TYPE_OLD_CHARGE;
}
if (input_type == TYPE_CONSENSUS)
{
// Every consensus style line includes features with four columns.
// The remainder is meta data
input_features = headers.size() / 4;
}
if (offset == 0 && (input_type == TYPE_OLD_CHARGE || input_type == TYPE_CONSENSUS))
{
throw Exception::ParseError(__FILE__, __LINE__, __PRETTY_FUNCTION__, "", String("Failed parsing in line 1: No HEADER provided. This is only allowed for three columns. You have more!\nOffending line: '") + header_trimmed + "' (line 1)\n");
}
SignedSize input_size = input.end() - input.begin();
ConsensusMap::FileDescription desc;
desc.filename = filename;
desc.size = (input_size) - offset;
consensus_map.getFileDescriptions()[0] = desc;
// parsing features
consensus_map.reserve(input_size);
for (; input_it != input.end(); ++input_it)
{
//do nothing for empty lines
String line_trimmed = *input_it;
line_trimmed.trim();
if (line_trimmed == "")
{
if ((input_it - input.begin()) < input_size - 1) LOG_WARN << "Notice: Empty line ignored (line " << ((input_it - input.begin()) + 1) << ").";
continue;
}
//split line to tokens
std::vector<String> parts;
input_it->split(separator, parts);
//abort if line does not contain enough fields
if (parts.size() < 3)
{
throw Exception::ParseError(__FILE__, __LINE__, __PRETTY_FUNCTION__, "",
String("Failed parsing in line ")
+ String((input_it - input.begin()) + 1)
+ ": At least three columns are needed! (got "
+ String(parts.size())
+ ")\nOffending line: '"
+ line_trimmed
+ "' (line "
+ String((input_it - input.begin()) + 1)
+ ")\n");
}
ConsensusFeature cf;
cf.setUniqueId();
try
{
// Convert values. Will return -1 if not available.
rt = checkedToDouble_(parts, 0);
mz = checkedToDouble_(parts, 1);
it = checkedToDouble_(parts, 2);
ch = checkedToInt_(parts, 3);
cf.setRT(rt);
cf.setMZ(mz);
cf.setIntensity(it);
if (input_type != TYPE_OLD_NOCHARGE)
cf.setCharge(ch);
}
catch (Exception::BaseException&)
{
throw Exception::ParseError(__FILE__, __LINE__, __PRETTY_FUNCTION__, "", String("Failed parsing in line ") + String((input_it - input.begin()) + 1) + ": Could not convert the first three columns to a number!\nOffending line: '" + line_trimmed + "' (line " + String((input_it - input.begin()) + 1) + ")\n");
}
// Check all features in one line
for (Size j = 1; j < input_features; ++j)
{
try
{
Feature f;
f.setUniqueId();
// Convert values. Will return -1 if not available.
rt = checkedToDouble_(parts, j * 4 + 0);
mz = checkedToDouble_(parts, j * 4 + 1);
it = checkedToDouble_(parts, j * 4 + 2);
ch = checkedToInt_(parts, j * 4 + 3);
// Only accept features with at least RT and MZ set
if (rt != -1 && mz != -1)
{
f.setRT(rt);
f.setMZ(mz);
f.setIntensity(it);
f.setCharge(ch);
cf.insert(j - 1, f);
}
}
catch (Exception::BaseException&)
{
throw Exception::ParseError(__FILE__, __LINE__, __PRETTY_FUNCTION__, "", String("Failed parsing in line ") + String((input_it - input.begin()) + 1) + ": Could not convert one of the four sub-feature columns (starting at column " + (j * 4 + 1) + ") to a number! Is the correct separator specified?\nOffending line: '" + line_trimmed + "' (line " + String((input_it - input.begin()) + 1) + ")\n");
}
}
//parse meta data
for (Size j = input_features * 4; j < parts.size(); ++j)
{
String part_trimmed = parts[j];
part_trimmed.trim();
if (part_trimmed != "")
{
//check if column name is ok
if (headers.size() <= j || headers[j] == "")
{
throw Exception::ParseError(__FILE__, __LINE__, __PRETTY_FUNCTION__, "",
String("Error: Missing meta data header for column ") + (j + 1) + "!"
+ String("Offending header line: '") + header_trimmed + "' (line 1)");
}
//add meta value
cf.setMetaValue(headers[j], part_trimmed);
}
}
//insert feature to map
consensus_map.push_back(cf);
}
// register FileDescriptions
ConsensusMap::FileDescription fd;
fd.filename = filename;
fd.size = consensus_map.size();
Size maps = std::max(input_features - 1, Size(1)); // its either a simple feature or a consensus map
// (in this case the 'input_features' includes the centroid, which we do not count)
for (Size i = 0; i < maps; ++i)
{
fd.label = String("EDTA_Map ") + String(i);
consensus_map.getFileDescriptions()[i] = fd;
}
}
/// @brief extracts the iTRAQ channels from the MS data and stores intensity values in a consensus map
///
/// @param ms_exp_data Raw data to read
/// @param consensus_map Output each MS² scan as a consensus feature
/// @throws Exception::MissingInformation if no scans present or MS² scan has no precursor
void ItraqChannelExtractor::run(const MSExperiment<Peak1D>& ms_exp_data, ConsensusMap& consensus_map)
{
if (ms_exp_data.empty())
{
LOG_WARN << "The given file does not contain any conventional peak data, but might"
" contain chromatograms. This tool currently cannot handle them, sorry.";
throw Exception::MissingInformation(__FILE__, __LINE__, __PRETTY_FUNCTION__, "Experiment has no scans!");
}
MSExperiment<> ms_exp_MS2;
String mode = (String) param_.getValue("select_activation");
std::cout << "Selecting scans with activation mode: " << (mode == "" ? "any" : mode) << "\n";
HasActivationMethod<MSExperiment<Peak1D>::SpectrumType> activation_predicate(ListUtils::create<String>(mode));
for (size_t idx = 0; idx < ms_exp_data.size(); ++idx)
{
if (ms_exp_data[idx].getMSLevel() == 2)
{
if (mode == "" || activation_predicate(ms_exp_data[idx]))
{
// copy only MS² scans
ms_exp_MS2.addSpectrum(ms_exp_data[idx]);
}
else
{
//std::cout << "deleting spectrum # " << idx << " with RT: " << ms_exp_data[idx].getRT() << "\n";
}
}
}
#ifdef ITRAQ_DEBUG
std::cout << "we have " << ms_exp_MS2.size() << " scans left of level " << ms_exp_MS2[0].getMSLevel() << std::endl;
std::cout << "run: channel_map_ has " << channel_map_.size() << " entries!" << std::endl;
#endif
consensus_map.clear(false);
// set <mapList> header
Int index_cnt = 0;
for (ChannelMapType::const_iterator cm_it = channel_map_.begin(); cm_it != channel_map_.end(); ++cm_it)
{
// structure of Map cm_it
// first == channel-name as Int e.g. 114
// second == ChannelInfo struct
ConsensusMap::FileDescription channel_as_map;
// label is the channel + description provided in the Params
if (itraq_type_ != TMT_SIXPLEX)
channel_as_map.label = "iTRAQ_" + String(cm_it->second.name) + "_" + String(cm_it->second.description);
else
channel_as_map.label = "TMT_" + String(cm_it->second.name) + "_" + String(cm_it->second.description);
channel_as_map.size = ms_exp_MS2.size();
//TODO what about .filename? leave empty?
// add some more MetaInfo
channel_as_map.setMetaValue("channel_name", cm_it->second.name);
channel_as_map.setMetaValue("channel_id", cm_it->second.id);
channel_as_map.setMetaValue("channel_description", cm_it->second.description);
channel_as_map.setMetaValue("channel_center", cm_it->second.center);
channel_as_map.setMetaValue("channel_active", String(cm_it->second.active ? "true" : "false"));
consensus_map.getFileDescriptions()[index_cnt++] = channel_as_map;
}
// create consensusElements
Peak2D::CoordinateType allowed_deviation = (Peak2D::CoordinateType) param_.getValue("reporter_mass_shift");
// now we have picked data
// --> assign peaks to channels
UInt element_index(0);
for (MSExperiment<>::ConstIterator it = ms_exp_MS2.begin(); it != ms_exp_MS2.end(); ++it)
{
// store RT&MZ of parent ion as centroid of ConsensusFeature
ConsensusFeature cf;
cf.setUniqueId();
cf.setRT(it->getRT());
if (it->getPrecursors().size() >= 1)
{
cf.setMZ(it->getPrecursors()[0].getMZ());
}
else
{
throw Exception::MissingInformation(__FILE__, __LINE__, __PRETTY_FUNCTION__, String("No precursor information given for scan native ID ") + String(it->getNativeID()) + " with RT " + String(it->getRT()));
}
Peak2D channel_value;
channel_value.setRT(it->getRT());
// for each each channel
Int index = 0;
Peak2D::IntensityType overall_intensity = 0;
for (ChannelMapType::const_iterator cm_it = channel_map_.begin(); cm_it != channel_map_.end(); ++cm_it)
{
// set mz-position of channel
channel_value.setMZ(cm_it->second.center);
// reset intensity
channel_value.setIntensity(0);
//add up all signals
for (MSExperiment<>::SpectrumType::ConstIterator mz_it =
it->MZBegin(cm_it->second.center - allowed_deviation)
; mz_it != it->MZEnd(cm_it->second.center + allowed_deviation)
; ++mz_it
)
{
channel_value.setIntensity(channel_value.getIntensity() + mz_it->getIntensity());
}
overall_intensity += channel_value.getIntensity();
// add channel to ConsensusFeature
cf.insert(index++, channel_value, element_index);
} // ! channel_iterator
// check featureHandles are not empty
if (overall_intensity == 0)
{
cf.setMetaValue("all_empty", String("true"));
}
cf.setIntensity(overall_intensity);
consensus_map.push_back(cf);
// the tandem-scan in the order they appear in the experiment
++element_index;
} // ! Experiment iterator
#ifdef ITRAQ_DEBUG
std::cout << "processed " << element_index << " scans" << std::endl;
#endif
consensus_map.setExperimentType("itraq");
return;
}
void QTClusterFinder::makeConsensusFeature_(list<QTCluster> & clustering,
ConsensusFeature & feature, OpenMSBoost::unordered_map<GridFeature *,
std::vector< QTCluster * > > & element_mapping)
{
// find the best cluster (a valid cluster with the highest score)
list<QTCluster>::iterator best = clustering.begin();
while (best != clustering.end() && best->isInvalid()) {++best;}
for (list<QTCluster>::iterator it = best;
it != clustering.end(); ++it)
{
if (!it->isInvalid())
{
if (it->getQuality() > best->getQuality())
{
best = it;
}
}
}
// no more clusters to process -> clear clustering and return
if (best == clustering.end())
{
clustering.clear();
return;
}
OpenMSBoost::unordered_map<Size, GridFeature *> elements;
best->getElements(elements);
// cout << "Elements: " << elements.size() << " with best " << best->getQuality() << " invalid " << best->isInvalid() << endl;
// create consensus feature from best cluster:
feature.setQuality(best->getQuality());
for (OpenMSBoost::unordered_map<Size, GridFeature *>::const_iterator it = elements.begin();
it != elements.end(); ++it)
{
feature.insert(it->first, it->second->getFeature());
}
feature.computeConsensus();
// update the clustering:
// 1. remove current "best" cluster
// 2. update all clusters accordingly and invalidate elements whose central
// element is removed
best->setInvalid();
for (OpenMSBoost::unordered_map<Size, GridFeature *>::const_iterator it = elements.begin();
it != elements.end(); ++it)
{
for (std::vector< QTCluster* >::iterator
cluster = element_mapping[&(*it->second)].begin();
cluster != element_mapping[&(*it->second)].end(); ++cluster)
{
// we do not want to update invalid features (saves time and does not
// recompute the quality)
if (!(*cluster)->isInvalid())
{
if (!(*cluster)->update(elements)) // cluster is invalid (center point removed):
{
(*cluster)->setInvalid();
}
}
}
}
}
ExitCodes main_(int, const char**)
{
//-------------------------------------------------------------
// parameter handling
//-------------------------------------------------------------
StringList in = getStringList_("in");
String edta = getStringOption_("pos");
String out = getStringOption_("out");
String out_sep = getStringOption_("out_separator");
String out_TIC_debug = getStringOption_("auto_rt:out_debug_TIC");
StringList in_header = getStringList_("in_header");
// number of out_debug_TIC files and input files must be identical
/*if (out_TIC_debug.size() > 0 && in.size() != out_TIC_debug.size())
{
LOG_FATAL_ERROR << "Error: number of input file 'in' and auto_rt:out_debug_TIC files must be identical!" << std::endl;
return ILLEGAL_PARAMETERS;
}*/
// number of header files and input files must be identical
if (in_header.size() > 0 && in.size() != in_header.size())
{
LOG_FATAL_ERROR << "Error: number of input file 'in' and 'in_header' files must be identical!" << std::endl;
return ILLEGAL_PARAMETERS;
}
if (!getFlag_("auto_rt:enabled") && !out_TIC_debug.empty())
{
LOG_FATAL_ERROR << "Error: TIC output file requested, but auto_rt is not enabled! Either do not request the file or switch on 'auto_rt:enabled'." << std::endl;
return ILLEGAL_PARAMETERS;
}
double rttol = getDoubleOption_("rt_tol");
double mztol = getDoubleOption_("mz_tol");
Size rt_collect = getIntOption_("rt_collect");
//-------------------------------------------------------------
// loading input
//-------------------------------------------------------------
MzMLFile mzml_file;
mzml_file.setLogType(log_type_);
MSExperiment<Peak1D> exp, exp_pp;
EDTAFile ed;
ConsensusMap cm;
ed.load(edta, cm);
StringList tf_single_header0, tf_single_header1, tf_single_header2; // header content, for each column
std::vector<String> vec_single; // one line for each compound, multiple columns per experiment
vec_single.resize(cm.size());
for (Size fi = 0; fi < in.size(); ++fi)
{
// load raw data
mzml_file.load(in[fi], exp);
exp.sortSpectra(true);
if (exp.empty())
{
LOG_WARN << "The given file does not contain any conventional peak data, but might"
" contain chromatograms. This tool currently cannot handle them, sorry." << std::endl;
return INCOMPATIBLE_INPUT_DATA;
}
// try to detect RT peaks (only for the first input file -- all others should align!)
// cm.size() might change in here...
if (getFlag_("auto_rt:enabled") && fi == 0)
{
ConsensusMap cm_local = cm; // we might have different RT peaks for each map if 'auto_rt' is enabled
cm.clear(false); // reset global list (about to be filled)
// compute TIC
MSChromatogram<> tic = exp.getTIC();
MSSpectrum<> tics, tic_gf, tics_pp, tics_sn;
for (Size ic = 0; ic < tic.size(); ++ic)
{ // rewrite Chromatogram to MSSpectrum (GaussFilter requires it)
Peak1D peak;
peak.setMZ(tic[ic].getRT());
peak.setIntensity(tic[ic].getIntensity());
tics.push_back(peak);
}
// smooth (no PP_CWT here due to efficiency reasons -- large FWHM take longer!)
double fwhm = getDoubleOption_("auto_rt:FHWM");
GaussFilter gf;
Param p = gf.getParameters();
p.setValue("gaussian_width", fwhm * 2); // wider than FWHM, just to be sure we have a fully smoothed peak. Merging two peaks is unlikely
p.setValue("use_ppm_tolerance", "false");
gf.setParameters(p);
tic_gf = tics;
gf.filter(tic_gf);
// pick peaks
PeakPickerHiRes pp;
p = pp.getParameters();
p.setValue("signal_to_noise", getDoubleOption_("auto_rt:SNThreshold"));
pp.setParameters(p);
pp.pick(tic_gf, tics_pp);
if (tics_pp.size())
{
LOG_INFO << "Found " << tics_pp.size() << " auto-rt peaks at: ";
for (Size ipp = 0; ipp != tics_pp.size(); ++ipp) LOG_INFO << " " << tics_pp[ipp].getMZ();
}
else
{
LOG_INFO << "Found no auto-rt peaks. Change threshold parameters!";
}
LOG_INFO << std::endl;
if (!out_TIC_debug.empty()) // if debug file was given
{ // store intermediate steps for debug
MSExperiment<> out_debug;
out_debug.addChromatogram(toChromatogram(tics));
out_debug.addChromatogram(toChromatogram(tic_gf));
SignalToNoiseEstimatorMedian<MSSpectrum<> > snt;
snt.init(tics);
for (Size is = 0; is < tics.size(); ++is)
{
Peak1D peak;
peak.setMZ(tic[is].getMZ());
peak.setIntensity(snt.getSignalToNoise(tics[is]));
tics_sn.push_back(peak);
}
out_debug.addChromatogram(toChromatogram(tics_sn));
out_debug.addChromatogram(toChromatogram(tics_pp));
// get rid of "native-id" missing warning
for (Size id = 0; id < out_debug.size(); ++id) out_debug[id].setNativeID(String("spectrum=") + id);
mzml_file.store(out_TIC_debug, out_debug);
LOG_DEBUG << "Storing debug AUTO-RT: " << out_TIC_debug << std::endl;
}
// add target EICs: for each m/z with no/negative RT, add all combinations of that m/z with auto-RTs
// duplicate m/z entries will be ignored!
// all other lines with positive RT values are copied unaffected
//do not allow doubles
std::set<double> mz_doubles;
for (ConsensusMap::Iterator cit = cm_local.begin(); cit != cm_local.end(); ++cit)
{
if (cit->getRT() < 0)
{
if (mz_doubles.find(cit->getMZ()) == mz_doubles.end())
{
mz_doubles.insert(cit->getMZ());
}
else
{
LOG_INFO << "Found duplicate m/z entry (" << cit->getMZ() << ") for auto-rt. Skipping ..." << std::endl;
continue;
}
ConsensusMap cm_RT_multiplex;
for (MSSpectrum<>::ConstIterator itp = tics_pp.begin(); itp != tics_pp.end(); ++itp)
{
ConsensusFeature f = *cit;
f.setRT(itp->getMZ());
cm.push_back(f);
}
}
else
{ // default feature with no auto-rt
LOG_INFO << "copying feature with RT " << cit->getRT() << std::endl;
cm.push_back(*cit);
}
}
// resize, since we have more positions now
vec_single.resize(cm.size());
}
// search for each EIC and add up
Int not_found(0);
Map<Size, double> quant;
String description;
if (fi < in_header.size())
{
HeaderInfo info(in_header[fi]);
description = info.header_description;
}
if (fi == 0)
{ // two additional columns for first file (theoretical RT and m/z)
tf_single_header0 << "" << "";
tf_single_header1 << "" << "";
tf_single_header2 << "RT" << "mz";
}
// 5 entries for each input file
tf_single_header0 << File::basename(in[fi]) << "" << "" << "" << "";
tf_single_header1 << description << "" << "" << "" << "";
tf_single_header2 << "RTobs" << "dRT" << "mzobs" << "dppm" << "intensity";
for (Size i = 0; i < cm.size(); ++i)
{
//std::cerr << "Rt" << cm[i].getRT() << " mz: " << cm[i].getMZ() << " R " << cm[i].getMetaValue("rank") << "\n";
double mz_da = mztol * cm[i].getMZ() / 1e6; // mz tolerance in Dalton
MSExperiment<>::ConstAreaIterator it = exp.areaBeginConst(cm[i].getRT() - rttol / 2,
cm[i].getRT() + rttol / 2,
cm[i].getMZ() - mz_da,
cm[i].getMZ() + mz_da);
Peak2D max_peak;
max_peak.setIntensity(0);
max_peak.setRT(cm[i].getRT());
max_peak.setMZ(cm[i].getMZ());
for (; it != exp.areaEndConst(); ++it)
{
if (max_peak.getIntensity() < it->getIntensity())
{
max_peak.setIntensity(it->getIntensity());
max_peak.setRT(it.getRT());
max_peak.setMZ(it->getMZ());
}
}
double ppm = 0; // observed m/z offset
if (max_peak.getIntensity() == 0)
{
++not_found;
}
else
{
// take median for m/z found
std::vector<double> mz;
MSExperiment<>::Iterator itm = exp.RTBegin(max_peak.getRT());
SignedSize low = std::min<SignedSize>(std::distance(exp.begin(), itm), rt_collect);
SignedSize high = std::min<SignedSize>(std::distance(itm, exp.end()) - 1, rt_collect);
MSExperiment<>::AreaIterator itt = exp.areaBegin((itm - low)->getRT() - 0.01, (itm + high)->getRT() + 0.01, cm[i].getMZ() - mz_da, cm[i].getMZ() + mz_da);
for (; itt != exp.areaEnd(); ++itt)
{
mz.push_back(itt->getMZ());
//std::cerr << "ppm: " << itt.getRT() << " " << itt->getMZ() << " " << itt->getIntensity() << std::endl;
}
if ((SignedSize)mz.size() > (low + high + 1)) LOG_WARN << "Compound " << i << " has overlapping peaks [" << mz.size() << "/" << low + high + 1 << "]" << std::endl;
if (!mz.empty())
{
double avg_mz = std::accumulate(mz.begin(), mz.end(), 0.0) / double(mz.size());
//std::cerr << "avg: " << avg_mz << "\n";
ppm = (avg_mz - cm[i].getMZ()) / cm[i].getMZ() * 1e6;
}
}
// appending the second column set requires separator
String append_sep = (fi == 0 ? "" : out_sep);
vec_single[i] += append_sep; // new line
if (fi == 0)
{
vec_single[i] += String(cm[i].getRT()) + out_sep +
String(cm[i].getMZ()) + out_sep;
}
vec_single[i] += String(max_peak.getRT()) + out_sep +
String(max_peak.getRT() - cm[i].getRT()) + out_sep +
String(max_peak.getMZ()) + out_sep +
String(ppm) + out_sep +
String(max_peak.getIntensity());
}
if (not_found) LOG_INFO << "Missing peaks for " << not_found << " compounds in file '" << in[fi] << "'.\n";
}
//-------------------------------------------------------------
// create header
//-------------------------------------------------------------
vec_single.insert(vec_single.begin(), ListUtils::concatenate(tf_single_header2, out_sep));
vec_single.insert(vec_single.begin(), ListUtils::concatenate(tf_single_header1, out_sep));
vec_single.insert(vec_single.begin(), ListUtils::concatenate(tf_single_header0, out_sep));
//-------------------------------------------------------------
// writing output
//-------------------------------------------------------------
TextFile tf;
for (std::vector<String>::iterator v_it = vec_single.begin(); v_it != vec_single.end(); ++v_it)
{
tf.addLine(*v_it);
}
tf.store(out);
return EXECUTION_OK;
}
void SILACLabeler::postDigestHook(SimTypes::FeatureMapSimVector& features_to_simulate)
{
SimTypes::FeatureMapSim& light_channel_features = features_to_simulate[0];
SimTypes::FeatureMapSim& medium_channel_features = features_to_simulate[1];
// merge the generated feature maps and create consensus
SimTypes::FeatureMapSim final_feature_map = mergeProteinIdentificationsMaps_(features_to_simulate);
if (features_to_simulate.size() == 2)
{
Map<String, Feature> unlabeled_features_index;
for (SimTypes::FeatureMapSim::iterator unlabeled_features_iter = light_channel_features.begin();
unlabeled_features_iter != light_channel_features.end();
++unlabeled_features_iter)
{
(*unlabeled_features_iter).ensureUniqueId();
unlabeled_features_index.insert(std::make_pair(
(*unlabeled_features_iter).getPeptideIdentifications()[0].getHits()[0].getSequence().toString()
,
*unlabeled_features_iter
));
}
// iterate over second map
for (SimTypes::FeatureMapSim::iterator labeled_feature_iter = medium_channel_features.begin(); labeled_feature_iter != medium_channel_features.end(); ++labeled_feature_iter)
{
const String unmodified_sequence = getUnmodifiedSequence_(*labeled_feature_iter, medium_channel_arginine_label_, medium_channel_lysine_label_);
// guarantee uniqueness
(*labeled_feature_iter).ensureUniqueId();
// check if we have a pair
if (unlabeled_features_index.has(unmodified_sequence))
{
// own scope as we don't know what happens to 'f_modified' once we call erase() below
Feature& unlabeled_feature = unlabeled_features_index[unmodified_sequence];
// guarantee uniqueness
unlabeled_feature.ensureUniqueId();
// feature has a SILAC Label and is not equal to non-labeled
if ((*labeled_feature_iter).getPeptideIdentifications()[0].getHits()[0].getSequence().isModified())
{
// add features to final map
final_feature_map.push_back(*labeled_feature_iter);
final_feature_map.push_back(unlabeled_feature);
// create consensus feature
ConsensusFeature cf;
cf.insert(MEDIUM_FEATURE_MAPID_, *labeled_feature_iter);
cf.insert(LIGHT_FEATURE_MAPID_, unlabeled_feature);
cf.ensureUniqueId();
consensus_.push_back(cf);
// remove unlabeled feature
unlabeled_features_index.erase(unmodified_sequence);
}
else
{
// merge features since they are equal
Feature final_feature = mergeFeatures_(*labeled_feature_iter, unmodified_sequence, unlabeled_features_index, 1, 2);
final_feature_map.push_back(final_feature);
}
}
else // no SILAC pair, just add the labeled one
{
final_feature_map.push_back(*labeled_feature_iter);
}
}
// add singletons from unlabeled channel
// clean up unlabeled_index
for (Map<String, Feature>::iterator unlabeled_index_iter = unlabeled_features_index.begin(); unlabeled_index_iter != unlabeled_features_index.end(); ++unlabeled_index_iter)
{
// the single ones from c0
final_feature_map.push_back(unlabeled_index_iter->second);
}
}
// merge three channels
if (features_to_simulate.size() == 3)
{
// index of unlabeled channelunlabeled_feature
Map<String, Feature> unlabeled_features_index;
for (SimTypes::FeatureMapSim::iterator unlabeled_features_iter = light_channel_features.begin();
unlabeled_features_iter != light_channel_features.end();
++unlabeled_features_iter)
{
(*unlabeled_features_iter).ensureUniqueId();
unlabeled_features_index.insert(std::make_pair(
(*unlabeled_features_iter).getPeptideIdentifications()[0].getHits()[0].getSequence().toString()
,
*unlabeled_features_iter
));
}
// index of labeled channel
Map<String, Feature> medium_features_index;
for (SimTypes::FeatureMapSim::iterator labeled_features_iter = medium_channel_features.begin();
labeled_features_iter != medium_channel_features.end();
++labeled_features_iter)
{
(*labeled_features_iter).ensureUniqueId();
medium_features_index.insert(std::make_pair(
getUnmodifiedSequence_(*labeled_features_iter, medium_channel_arginine_label_, medium_channel_lysine_label_)
,
*labeled_features_iter
));
}
SimTypes::FeatureMapSim& heavy_labeled_features = features_to_simulate[2];
for (SimTypes::FeatureMapSim::iterator heavy_labeled_feature_iter = heavy_labeled_features.begin();
heavy_labeled_feature_iter != heavy_labeled_features.end();
++heavy_labeled_feature_iter)
{
Feature& heavy_feature = *heavy_labeled_feature_iter;
heavy_feature.ensureUniqueId();
String heavy_feature_unmodified_sequence = getUnmodifiedSequence_(heavy_feature, heavy_channel_arginine_label_, heavy_channel_lysine_label_);
if (unlabeled_features_index.has(heavy_feature_unmodified_sequence) && medium_features_index.has(heavy_feature_unmodified_sequence))
{
// it is a triplet
// c2 & c1 modified
if (heavy_feature_unmodified_sequence != heavy_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().toString())
{
// add features to final map
final_feature_map.push_back(heavy_feature);
final_feature_map.push_back(medium_features_index[heavy_feature_unmodified_sequence]);
final_feature_map.push_back(unlabeled_features_index[heavy_feature_unmodified_sequence]);
ConsensusFeature c_triplet;
c_triplet.insert(HEAVY_FEATURE_MAPID_, heavy_feature);
c_triplet.insert(LIGHT_FEATURE_MAPID_, unlabeled_features_index[heavy_feature_unmodified_sequence]);
c_triplet.insert(MEDIUM_FEATURE_MAPID_, medium_features_index[heavy_feature_unmodified_sequence]);
c_triplet.ensureUniqueId();
consensus_.push_back(c_triplet);
}
else
{
// merge all three channels
Feature completeMerge = mergeAllChannelFeatures_(heavy_feature, heavy_feature_unmodified_sequence, unlabeled_features_index, medium_features_index);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
unlabeled_features_index.erase(heavy_feature_unmodified_sequence);
medium_features_index.erase(heavy_feature_unmodified_sequence);
}
else if (unlabeled_features_index.has(heavy_feature_unmodified_sequence))
{
// 2nd case light and heavy pair
if (heavy_feature_unmodified_sequence != heavy_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().toString())
{
// add features to final map
final_feature_map.push_back(heavy_feature);
final_feature_map.push_back(unlabeled_features_index[heavy_feature_unmodified_sequence]);
ConsensusFeature c_duplet;
c_duplet.insert(HEAVY_FEATURE_MAPID_, heavy_feature);
c_duplet.insert(LIGHT_FEATURE_MAPID_, unlabeled_features_index[heavy_feature_unmodified_sequence]);
c_duplet.ensureUniqueId();
consensus_.push_back(c_duplet);
}
else
{
// merge all three channels
Feature completeMerge = mergeFeatures_(heavy_feature, heavy_feature_unmodified_sequence, unlabeled_features_index, 1, 3);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
unlabeled_features_index.erase(heavy_feature_unmodified_sequence);
}
else if (medium_features_index.has(heavy_feature_unmodified_sequence))
{
// 3rd case medium and heavy pair
if (heavy_feature_unmodified_sequence != heavy_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().toString())
{
// add features to final map
final_feature_map.push_back(heavy_feature);
final_feature_map.push_back(medium_features_index[heavy_feature_unmodified_sequence]);
ConsensusFeature c_duplet;
c_duplet.insert(HEAVY_FEATURE_MAPID_, heavy_feature);
c_duplet.insert(MEDIUM_FEATURE_MAPID_, medium_features_index[heavy_feature_unmodified_sequence]);
c_duplet.ensureUniqueId();
consensus_.push_back(c_duplet);
}
else
{
// merge all
Feature completeMerge = mergeFeatures_(heavy_feature, heavy_feature_unmodified_sequence, medium_features_index, 2, 3);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
medium_features_index.erase(heavy_feature_unmodified_sequence);
}
else
{
// heavy feature is a singleton
final_feature_map.push_back(heavy_feature);
}
}
// clean up labeled_index
for (Map<String, Feature>::iterator medium_channle_index_iterator = medium_features_index.begin(); medium_channle_index_iterator != medium_features_index.end(); ++medium_channle_index_iterator)
{
Feature& medium_channel_feature = medium_channle_index_iterator->second;
medium_channel_feature.ensureUniqueId();
String medium_channel_feature_unmodified_sequence = getUnmodifiedSequence_(medium_channel_feature, medium_channel_arginine_label_, medium_channel_lysine_label_);
if (unlabeled_features_index.has(medium_channel_feature_unmodified_sequence))
{
// 1. case: pair between c0 and c1
if (medium_channel_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().isModified())
{
// add features to final map
final_feature_map.push_back(medium_channel_feature);
final_feature_map.push_back(unlabeled_features_index[medium_channel_feature_unmodified_sequence]);
ConsensusFeature c_duplet;
c_duplet.insert(MEDIUM_FEATURE_MAPID_, medium_channel_feature);
c_duplet.insert(LIGHT_FEATURE_MAPID_, unlabeled_features_index[medium_channel_feature_unmodified_sequence]);
c_duplet.ensureUniqueId();
consensus_.push_back(c_duplet);
}
else
{
// merge
Feature completeMerge = mergeFeatures_(medium_channel_feature, medium_channel_feature_unmodified_sequence, unlabeled_features_index, 1, 2);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
unlabeled_features_index.erase(medium_channel_feature_unmodified_sequence);
}
else
{
// c1 is alone
final_feature_map.push_back(medium_channel_feature);
}
}
// clean up unlabeled_index
for (Map<String, Feature>::iterator unlabeled_index_iter = unlabeled_features_index.begin(); unlabeled_index_iter != unlabeled_features_index.end(); ++unlabeled_index_iter)
{
// the single ones from c0
final_feature_map.push_back(unlabeled_index_iter->second);
}
}
features_to_simulate.clear();
features_to_simulate.push_back(final_feature_map);
consensus_.setProteinIdentifications(final_feature_map.getProteinIdentifications());
ConsensusMap::FileDescription map_description;
map_description.label = "Simulation (Labeling Consensus)";
map_description.size = features_to_simulate.size();
consensus_.getFileDescriptions()[0] = map_description;
}
void BaseLabeler::recomputeConsensus_(const FeatureMapSim & simulated_features)
{
// iterate over all given features stored in the labeling consensus and try to find the corresponding feature in
// in the feature map
// build index for faster access
Map<String, IntList> id_map;
Map<UInt64, Size> features_per_labeled_map;
for (Size i = 0; i < simulated_features.size(); ++i)
{
if (simulated_features[i].metaValueExists("parent_feature"))
{
LOG_DEBUG << "Checking [" << i << "]: " << simulated_features[i].getPeptideIdentifications()[0].getHits()[0].getSequence().toString()
<< " with charge " << simulated_features[i].getCharge() << " (" << simulated_features[i].getMetaValue("charge_adducts") << ")"
<< " parent was " << simulated_features[i].getMetaValue("parent_feature") << std::endl;
id_map[simulated_features[i].getMetaValue("parent_feature")].push_back((Int)i);
UInt64 map_index = 0;
if (simulated_features[i].metaValueExists("map_index"))
{
map_index = simulated_features[i].getMetaValue("map_index");
}
++features_per_labeled_map[map_index];
}
}
for (Map<String, IntList>::iterator it = id_map.begin(); it != id_map.end(); ++it)
{
LOG_DEBUG << it->first << " " << it->second << std::endl;
}
// new consensus map
ConsensusMap new_cm;
// initialize submaps in consensus map
for (Map<UInt64, Size>::Iterator it = features_per_labeled_map.begin(); it != features_per_labeled_map.end(); ++it)
{
new_cm.getFileDescriptions()[it->first].size = it->second;
new_cm.getFileDescriptions()[it->first].unique_id = simulated_features.getUniqueId();
}
for (ConsensusMap::iterator cm_iter = consensus_.begin(); cm_iter != consensus_.end(); ++cm_iter)
{
bool complete = true;
LOG_DEBUG << "Checking consensus feature containing: " << std::endl;
// check if we have all elements of current CF in the new feature map (simulated_features)
for (ConsensusFeature::iterator cf_iter = (*cm_iter).begin(); cf_iter != (*cm_iter).end(); ++cf_iter)
{
complete &= id_map.has(String((*cf_iter).getUniqueId()));
LOG_DEBUG << "\t" << String((*cf_iter).getUniqueId()) << std::endl;
}
if (complete)
{
// get all elements sorted by charge state; since the same charge can be achieved by different
// adduct compositions we use the adduct-string as indicator to find the groups
Map<String, std::set<FeatureHandle, FeatureHandle::IndexLess> > charge_mapping;
for (ConsensusFeature::iterator cf_iter = (*cm_iter).begin(); cf_iter != (*cm_iter).end(); ++cf_iter)
{
IntList feature_indices = id_map[String((*cf_iter).getUniqueId())];
for (IntList::iterator it = feature_indices.begin(); it != feature_indices.end(); ++it)
{
UInt64 map_index = 0;
if (simulated_features[*it].metaValueExists("map_index"))
{
map_index = simulated_features[*it].getMetaValue("map_index");
}
if (charge_mapping.has(simulated_features[*it].getMetaValue("charge_adducts")))
{
charge_mapping[simulated_features[*it].getMetaValue("charge_adducts")].insert(FeatureHandle(map_index, simulated_features[*it]));
}
else
{
LOG_DEBUG << "Create new set with charge composition " << simulated_features[*it].getMetaValue("charge_adducts") << std::endl;
std::set<FeatureHandle, FeatureHandle::IndexLess> fh_set;
fh_set.insert(FeatureHandle(map_index, simulated_features[*it]));
charge_mapping.insert(std::make_pair(simulated_features[*it].getMetaValue("charge_adducts"), fh_set));
}
}
}
// create new consensus feature from derived features (separated by charge, if charge != 0)
for (Map<String, std::set<FeatureHandle, FeatureHandle::IndexLess> >::const_iterator charge_group_it = charge_mapping.begin();
charge_group_it != charge_mapping.end();
++charge_group_it)
{
ConsensusFeature cf;
cf.setCharge((*(*charge_group_it).second.begin()).getCharge());
cf.setMetaValue("charge_adducts", charge_group_it->first);
std::vector<PeptideIdentification> ids;
for (std::set<FeatureHandle, FeatureHandle::IndexLess>::const_iterator fh_it = (charge_group_it->second).begin(); fh_it != (charge_group_it->second).end(); ++fh_it)
{
cf.insert(*fh_it);
// append identifications
Size f_index = simulated_features.uniqueIdToIndex(fh_it->getUniqueId());
std::vector<PeptideIdentification> ids_feature = simulated_features[f_index].getPeptideIdentifications();
ids.insert(ids.end(), ids_feature.begin(), ids_feature.end());
}
cf.computeMonoisotopicConsensus();
cf.setPeptideIdentifications(ids);
new_cm.push_back(cf);
}
}
}
new_cm.setProteinIdentifications(simulated_features.getProteinIdentifications());
consensus_.swap(new_cm);
consensus_.applyMemberFunction(&UniqueIdInterface::ensureUniqueId);
}
c2.insert(4,tmp_feature2); c2.insert(5,tmp_feature3); ConsensusFeature::MapsLess ml; TEST_EQUAL(ml(c1,c1), 0); TEST_EQUAL(ml(c1,c2), 1); TEST_EQUAL(ml(c2,c1), 0); TEST_EQUAL(ml(c2,c2), 0); END_SECTION START_SECTION((ConsensusFeature& operator=(const ConsensusFeature &rhs))) ConsensusFeature cons(tmp_feature); cons.insert(1,tmp_feature); ConsensusFeature cons_copy; cons_copy = cons; TEST_REAL_SIMILAR(cons_copy.getRT(),1.0) TEST_REAL_SIMILAR(cons_copy.getMZ(),2.0) TEST_REAL_SIMILAR(cons_copy.getIntensity(),200.0) TEST_EQUAL((cons_copy.begin())->getMapIndex(),1) TEST_EQUAL((cons_copy.begin())->getUniqueId(),3) TEST_EQUAL((cons_copy.begin())->getIntensity(),200) END_SECTION START_SECTION((ConsensusFeature(const ConsensusFeature &rhs))) ConsensusFeature cons(tmp_feature); cons.insert(1,tmp_feature); ConsensusFeature cons_copy(cons);
ExitCodes main_(int, const char**) override
{
//-------------------------------------------------------------
// parameter handling
//-------------------------------------------------------------
String in = getStringOption_("in");
String out = getStringOption_("out");
FileTypes::Type out_type = FileTypes::nameToType(getStringOption_("out_type"));
if (out_type == FileTypes::UNKNOWN)
{
out_type = FileHandler().getTypeByFileName(out);
}
//-------------------------------------------------------------
// loading input
//-------------------------------------------------------------
MzMLFile mz_data_file;
mz_data_file.setLogType(log_type_);
PeakMap ms_peakmap;
std::vector<Int> ms_level(1, 1);
(mz_data_file.getOptions()).setMSLevels(ms_level);
mz_data_file.load(in, ms_peakmap);
if (ms_peakmap.size() == 0)
{
LOG_WARN << "The given file does not contain any conventional peak data, but might"
" contain chromatograms. This tool currently cannot handle them, sorry.";
return INCOMPATIBLE_INPUT_DATA;
}
// make sure that the spectra are sorted by m/z
ms_peakmap.sortSpectra(true);
//-------------------------------------------------------------
// get params for MTD and EPD algorithms
//-------------------------------------------------------------
Param com_param = getParam_().copy("algorithm:common:", true);
writeDebug_("Common parameters passed to both sub-algorithms (mtd and epd)", com_param, 3);
Param mtd_param = getParam_().copy("algorithm:mtd:", true);
writeDebug_("Parameters passed to MassTraceDetection", mtd_param, 3);
Param epd_param = getParam_().copy("algorithm:epd:", true);
writeDebug_("Parameters passed to ElutionPeakDetection", epd_param, 3);
//-------------------------------------------------------------
// configure and run MTD
//-------------------------------------------------------------
MassTraceDetection mt_ext;
mtd_param.insert("", com_param);
mtd_param.remove("chrom_fwhm");
mt_ext.setParameters(mtd_param);
vector<MassTrace> m_traces;
mt_ext.run(ms_peakmap, m_traces);
vector<MassTrace> m_traces_final;
bool use_epd = epd_param.getValue("enabled").toBool();
if (!use_epd)
{
swap(m_traces_final, m_traces);
}
else
{
ElutionPeakDetection ep_det;
epd_param.remove("enabled"); // artificially added above
epd_param.insert("", com_param);
ep_det.setParameters(epd_param);
std::vector<MassTrace> split_mtraces;
// note: this step will destroy any meta data annotation (e.g. FWHM_mz_avg)
ep_det.detectPeaks(m_traces, split_mtraces);
if (ep_det.getParameters().getValue("width_filtering") == "auto")
{
m_traces_final.clear();
ep_det.filterByPeakWidth(split_mtraces, m_traces_final);
LOG_INFO << "Notice: " << split_mtraces.size() - m_traces_final.size()
<< " of total " << split_mtraces.size()
<< " were dropped because of too low peak width." << std::endl;
}
else
{
swap(m_traces_final, split_mtraces);
}
}
//-------------------------------------------------------------
// writing consensus map output
//-------------------------------------------------------------
if (out_type == FileTypes::CONSENSUSXML)
{
ConsensusMap consensus_map;
StringList ms_runs;
ms_peakmap.getPrimaryMSRunPath(ms_runs);
consensus_map.setPrimaryMSRunPath(ms_runs);
for (Size i = 0; i < m_traces_final.size(); ++i)
{
if (m_traces_final[i].getSize() == 0) continue;
ConsensusFeature fcons;
int k = 0;
for (MassTrace::const_iterator it = m_traces_final[i].begin(); it != m_traces_final[i].end(); ++it)
{
FeatureHandle fhandle;
fhandle.setRT(it->getRT());
fhandle.setMZ(it->getMZ());
fhandle.setIntensity(it->getIntensity());
fhandle.setUniqueId(++k);
fcons.insert(fhandle);
}
fcons.setMetaValue(3, m_traces_final[i].getLabel());
fcons.setCharge(0);
fcons.setWidth(m_traces_final[i].estimateFWHM(use_epd));
fcons.setQuality(1 - (1.0 / m_traces_final[i].getSize()));
fcons.setRT(m_traces_final[i].getCentroidRT());
fcons.setMZ(m_traces_final[i].getCentroidMZ());
fcons.setIntensity(m_traces_final[i].getIntensity(false));
consensus_map.push_back(fcons);
}
consensus_map.applyMemberFunction(&UniqueIdInterface::setUniqueId);
addDataProcessing_(consensus_map, getProcessingInfo_(DataProcessing::QUANTITATION));
consensus_map.setUniqueId();
ConsensusXMLFile().store(out, consensus_map);
}
else //(out_type == FileTypes::FEATUREXML)
{
//-----------------------------------------------------------
// convert mass traces to features
//-----------------------------------------------------------
std::vector<double> stats_sd;
FeatureMap ms_feat_map;
StringList ms_runs;
ms_peakmap.getPrimaryMSRunPath(ms_runs);
ms_feat_map.setPrimaryMSRunPath(ms_runs);
for (Size i = 0; i < m_traces_final.size(); ++i)
{
if (m_traces_final[i].getSize() == 0) continue;
m_traces_final[i].updateMeanMZ();
m_traces_final[i].updateWeightedMZsd();
Feature f;
f.setMetaValue(3, m_traces_final[i].getLabel());
f.setCharge(0);
f.setMZ(m_traces_final[i].getCentroidMZ());
f.setIntensity(m_traces_final[i].getIntensity(false));
f.setRT(m_traces_final[i].getCentroidRT());
f.setWidth(m_traces_final[i].estimateFWHM(use_epd));
f.setOverallQuality(1 - (1.0 / m_traces_final[i].getSize()));
f.getConvexHulls().push_back(m_traces_final[i].getConvexhull());
double sd = m_traces_final[i].getCentroidSD();
f.setMetaValue("SD", sd);
f.setMetaValue("SD_ppm", sd / f.getMZ() * 1e6);
if (m_traces_final[i].fwhm_mz_avg > 0) f.setMetaValue("FWHM_mz_avg", m_traces_final[i].fwhm_mz_avg);
stats_sd.push_back(m_traces_final[i].getCentroidSD());
ms_feat_map.push_back(f);
}
// print some stats about standard deviation of mass traces
if (stats_sd.size() > 0)
{
std::sort(stats_sd.begin(), stats_sd.end());
LOG_INFO << "Mass trace m/z s.d.\n"
<< " low quartile: " << stats_sd[stats_sd.size() * 1 / 4] << "\n"
<< " median: " << stats_sd[stats_sd.size() * 1 / 2] << "\n"
<< " upp quartile: " << stats_sd[stats_sd.size() * 3 / 4] << std::endl;
}
ms_feat_map.applyMemberFunction(&UniqueIdInterface::setUniqueId);
//-------------------------------------------------------------
// writing output
//-------------------------------------------------------------
// annotate output with data processing info TODO
addDataProcessing_(ms_feat_map, getProcessingInfo_(DataProcessing::QUANTITATION));
//ms_feat_map.setUniqueId();
FeatureXMLFile().store(out, ms_feat_map);
}
return EXECUTION_OK;
}
void ICPLLabeler::postDigestHook(SimTypes::FeatureMapSimVector& features_to_simulate)
{
SimTypes::FeatureMapSim& light_labeled_features = features_to_simulate[0];
SimTypes::FeatureMapSim& medium_labeled_features = features_to_simulate[1];
if (param_.getValue("label_proteins") == "false") // loop for peptide-labeling (post-digest-labeling)
{
// iterate over first map for light labeling
for (SimTypes::FeatureMapSim::iterator lf_iter = light_labeled_features.begin(); lf_iter != light_labeled_features.end(); ++lf_iter)
{
lf_iter->ensureUniqueId();
addModificationToPeptideHit_(*lf_iter, light_channel_label_);
}
// iterate over second map for medium labeling
for (SimTypes::FeatureMapSim::iterator lf_iter = medium_labeled_features.begin(); lf_iter != medium_labeled_features.end(); ++lf_iter)
{
lf_iter->ensureUniqueId();
addModificationToPeptideHit_(*lf_iter, medium_channel_label_);
}
if (features_to_simulate.size() == 3) //third channel labeling can only be done, if a third channel exist
{
SimTypes::FeatureMapSim& heavy_labeled_features = features_to_simulate[2];
// iterate over third map
for (SimTypes::FeatureMapSim::iterator lf_iter = heavy_labeled_features.begin(); lf_iter != heavy_labeled_features.end(); ++lf_iter)
{
lf_iter->ensureUniqueId();
addModificationToPeptideHit_(*lf_iter, heavy_channel_label_);
}
}
}
// merge the generated feature maps and create consensus
SimTypes::FeatureMapSim final_feature_map = mergeProteinIdentificationsMaps_(features_to_simulate);
if (features_to_simulate.size() == 2) // merge_modus for two FeatureMaps
{
// create index of light channel features for easy mapping of medium-to-light channel
Map<String, Feature> light_labeled_features_index;
for (SimTypes::FeatureMapSim::iterator light_labeled_features_iter = light_labeled_features.begin();
light_labeled_features_iter != light_labeled_features.end();
++light_labeled_features_iter)
{
(*light_labeled_features_iter).ensureUniqueId();
light_labeled_features_index.insert(std::make_pair(
getUnmodifiedAASequence_((*light_labeled_features_iter), light_channel_label_),
*light_labeled_features_iter
));
}
// iterate over second map
for (SimTypes::FeatureMapSim::iterator medium_labeled_feature_iter = medium_labeled_features.begin(); medium_labeled_feature_iter != medium_labeled_features.end(); ++medium_labeled_feature_iter)
{
AASequence medium_labeled_feature_sequence = (*medium_labeled_feature_iter).getPeptideIdentifications()[0].getHits()[0].getSequence();
// guarantee uniqueness
(*medium_labeled_feature_iter).ensureUniqueId();
// check if we have a pair
if (light_labeled_features_index.has(getUnmodifiedAASequence_((*medium_labeled_feature_iter), medium_channel_label_)))
{
// own scope as we don't know what happens to 'f_modified' once we call erase() below
Feature& light_labeled_feature = light_labeled_features_index[getUnmodifiedAASequence_((*medium_labeled_feature_iter), medium_channel_label_)];
// guarantee uniqueness
light_labeled_feature.ensureUniqueId();
if (medium_labeled_feature_sequence.isModified()) // feature has a medium ICPL-Label and is not equal to light-labeled
{
// add features to final map
final_feature_map.push_back(*medium_labeled_feature_iter);
final_feature_map.push_back(light_labeled_feature);
// create consensus feature
ConsensusFeature cf;
cf.insert(MEDIUM_FEATURE_MAPID_, *medium_labeled_feature_iter);
cf.insert(LIGHT_FEATURE_MAPID_, light_labeled_feature);
consensus_.push_back(cf);
// remove light-labeled feature
light_labeled_features_index.erase(getUnmodifiedAASequence_((*medium_labeled_feature_iter), medium_channel_label_));
}
else
{
// merge features since they are equal
Feature final_feature = mergeFeatures_(*medium_labeled_feature_iter, medium_labeled_feature_sequence, light_labeled_features_index);
final_feature_map.push_back(final_feature);
}
}
else // no ICPL pair, just add the medium-labeled one
{
final_feature_map.push_back(*medium_labeled_feature_iter);
}
}
// add singletons from light-labeled channel
// clean up light-labeled_index
for (Map<String, Feature>::iterator light_labeled_index_iter = light_labeled_features_index.begin(); light_labeled_index_iter != light_labeled_features_index.end(); ++light_labeled_index_iter)
{
// the single ones from c0
final_feature_map.push_back(light_labeled_index_iter->second);
}
}
else if (features_to_simulate.size() == 3) // merge_modus for three Channels
{
// create index of light channel features for easy mapping of heavy-to-medium-to-light channel
Map<String, Feature> light_labeled_features_index;
for (SimTypes::FeatureMapSim::iterator light_labeled_features_iter = light_labeled_features.begin();
light_labeled_features_iter != light_labeled_features.end();
++light_labeled_features_iter)
{
(*light_labeled_features_iter).ensureUniqueId();
light_labeled_features_index.insert(std::make_pair(
getUnmodifiedAASequence_(*light_labeled_features_iter, light_channel_label_),
*light_labeled_features_iter
));
}
// create index of medium channel features for easy mapping of heavy-to-medium-to-light channel
Map<String, Feature> medium_labeled_features_index;
for (SimTypes::FeatureMapSim::iterator medium_labeled_features_iter = medium_labeled_features.begin();
medium_labeled_features_iter != medium_labeled_features.end();
++medium_labeled_features_iter)
{
(*medium_labeled_features_iter).ensureUniqueId();
medium_labeled_features_index.insert(std::make_pair(
getUnmodifiedAASequence_((*medium_labeled_features_iter), medium_channel_label_),
*medium_labeled_features_iter
));
}
for (SimTypes::FeatureMapSim::iterator heavy_labeled_feature_iter = features_to_simulate[2].begin(); heavy_labeled_feature_iter != features_to_simulate[2].end(); ++heavy_labeled_feature_iter)
{
Feature& heavy_feature = *heavy_labeled_feature_iter;
String heavy_feature_unmodified_sequence = getUnmodifiedAASequence_(heavy_feature, heavy_channel_label_);
heavy_feature.ensureUniqueId();
if (light_labeled_features_index.has(heavy_feature_unmodified_sequence) && medium_labeled_features_index.has(heavy_feature_unmodified_sequence))
{
// 1st case .. it is a triplet
if (heavy_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().isModified())
{
// if heavy feature is modified, all light and medium channel are also
// add features to final map
final_feature_map.push_back(heavy_feature);
final_feature_map.push_back(medium_labeled_features_index[heavy_feature_unmodified_sequence]);
final_feature_map.push_back(light_labeled_features_index[heavy_feature_unmodified_sequence]);
// create triplet consensus feature
ConsensusFeature c_triplet;
c_triplet.insert(HEAVY_FEATURE_MAPID_, heavy_feature);
c_triplet.insert(LIGHT_FEATURE_MAPID_, light_labeled_features_index[heavy_feature_unmodified_sequence]);
c_triplet.insert(MEDIUM_FEATURE_MAPID_, medium_labeled_features_index[heavy_feature_unmodified_sequence]);
consensus_.push_back(c_triplet);
}
else
{
// merge all three channels
Feature c2c1 = mergeFeatures_(heavy_feature, AASequence::fromString(heavy_feature_unmodified_sequence), medium_labeled_features_index);
Feature completeMerge = mergeFeatures_(c2c1, AASequence::fromString(heavy_feature_unmodified_sequence), light_labeled_features_index);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
light_labeled_features_index.erase(heavy_feature_unmodified_sequence);
medium_labeled_features_index.erase(heavy_feature_unmodified_sequence);
}
else if (light_labeled_features_index.has(heavy_feature_unmodified_sequence))
{
// 2.Fall -> c0 - c2
if (heavy_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().isModified())
{
// add features to final map
final_feature_map.push_back(heavy_feature);
final_feature_map.push_back(light_labeled_features_index[heavy_feature_unmodified_sequence]);
ConsensusFeature c_triplet;
c_triplet.insert(HEAVY_FEATURE_MAPID_, heavy_feature);
c_triplet.insert(LIGHT_FEATURE_MAPID_, light_labeled_features_index[heavy_feature_unmodified_sequence]);
consensus_.push_back(c_triplet);
}
else
{
// merge all three channels
Feature completeMerge = mergeFeatures_(heavy_feature, AASequence::fromString(heavy_feature_unmodified_sequence), light_labeled_features_index);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
light_labeled_features_index.erase(heavy_feature_unmodified_sequence);
}
else if (medium_labeled_features_index.has(heavy_feature_unmodified_sequence))
{
// 3.Fall -> c1 - c2
if (heavy_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().isModified())
{
// add features to final map
final_feature_map.push_back(heavy_feature);
final_feature_map.push_back(medium_labeled_features_index[heavy_feature_unmodified_sequence]);
ConsensusFeature c_triplet;
c_triplet.insert(HEAVY_FEATURE_MAPID_, heavy_feature);
c_triplet.insert(MEDIUM_FEATURE_MAPID_, medium_labeled_features_index[heavy_feature_unmodified_sequence]);
consensus_.push_back(c_triplet);
}
else
{
// merge all
Feature completeMerge = mergeFeatures_(heavy_feature, AASequence::fromString(heavy_feature_unmodified_sequence), medium_labeled_features_index);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
medium_labeled_features_index.erase(heavy_feature_unmodified_sequence);
}
else
{
// 4.Fall -> alleine
final_feature_map.push_back(heavy_feature);
}
}
// clean up medium-labeled_index
for (Map<String, Feature>::iterator medium_labeled_index_iter = medium_labeled_features_index.begin(); medium_labeled_index_iter != medium_labeled_features_index.end(); ++medium_labeled_index_iter)
{
Feature& medium_labeled_feature = medium_labeled_index_iter->second;
medium_labeled_feature.ensureUniqueId();
String medium_labeled_feature_unmodified_sequence = medium_labeled_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().toUnmodifiedString();
if (light_labeled_features_index.has(medium_labeled_feature_unmodified_sequence))
{
// 1. case: pair between c0 and c1
if (medium_labeled_feature.getPeptideIdentifications()[0].getHits()[0].getSequence().isModified())
{
// add features to final map
final_feature_map.push_back(medium_labeled_feature);
final_feature_map.push_back(light_labeled_features_index[medium_labeled_feature_unmodified_sequence]);
ConsensusFeature c_triplet;
c_triplet.insert(MEDIUM_FEATURE_MAPID_, medium_labeled_feature);
c_triplet.insert(LIGHT_FEATURE_MAPID_, light_labeled_features_index[medium_labeled_feature_unmodified_sequence]);
consensus_.push_back(c_triplet);
}
else
{
// merge
Feature completeMerge = mergeFeatures_(medium_labeled_feature, AASequence::fromString(medium_labeled_feature_unmodified_sequence), light_labeled_features_index);
final_feature_map.push_back(completeMerge);
}
// remove features from indices
light_labeled_features_index.erase(medium_labeled_feature_unmodified_sequence);
}
else
{
// c1 is alone
final_feature_map.push_back(medium_labeled_feature);
}
}
// clean up light-labeled_index
for (Map<String, Feature>::iterator light_labeled_index_iter = light_labeled_features_index.begin(); light_labeled_index_iter != light_labeled_features_index.end(); ++light_labeled_index_iter)
{
// the single ones from c0
final_feature_map.push_back(light_labeled_index_iter->second);
}
}
features_to_simulate.clear();
features_to_simulate.push_back(final_feature_map);
consensus_.setProteinIdentifications(final_feature_map.getProteinIdentifications());
ConsensusMap::FileDescription map_description;
map_description.label = "Simulation (Labeling Consensus)";
map_description.size = features_to_simulate.size();
consensus_.getFileDescriptions()[0] = map_description;
}
