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;
}
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;
}
/// @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 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);
}
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;
}
}