void EDTAFile::store(const String& filename, const FeatureMap& map) const
{
TextFile tf;
tf.addLine("RT\tm/z\tintensity\tcharge");
for (Size i = 0; i < map.size(); ++i)
{
const Feature& f = map[i];
tf.addLine(String(f.getRT()) + "\t" + f.getMZ() + "\t" + f.getIntensity() + "\t" + f.getCharge());
}
tf.store(filename);
}
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);
}
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;
}
ExitCodes main_(int, const char**) override
{
//load data
FeatureMap features_in, features_truth;
FeatureXMLFile().load(getStringOption_("in"), features_in);
features_in.sortByPosition();
FeatureXMLFile().load(getStringOption_("truth"), features_truth);
features_truth.sortByPosition();
FeatureMap abort_reasons;
if (getStringOption_("abort_reasons") != "")
{
FeatureXMLFile().load(getStringOption_("abort_reasons"), abort_reasons);
}
double mz_tol = getDoubleOption_("mz_tol");
writeDebug_(String("Final MZ tolerance: ") + mz_tol, 1);
//determine average RT tolerance:
//median feature RT span times given factor
vector<double> rt_spans;
for (Size t = 0; t < features_in.size(); ++t)
{
if (features_in[t].getConvexHulls().size() != 0)
{
rt_spans.push_back(features_in[t].getConvexHull().getBoundingBox().width());
}
}
//feature convex hulls are available => relative RT span
double rt_tol = getDoubleOption_("rt_tol_abs");
if (rt_tol < 0.0)
{
if (!rt_spans.empty())
{
sort(rt_spans.begin(), rt_spans.end());
rt_tol = getDoubleOption_("rt_tol") * rt_spans[rt_spans.size() / 2];
}
else if (features_in.empty())
{
// do nothing, rt_tol does not really matter, as we will not find a match anyway, but we want to have the stats
// at the end, so we do not abort
}
else
{
writeLog_("Error: Input features do not have convex hulls. You have to set 'rt_tol_abs'!");
return ILLEGAL_PARAMETERS;
}
}
writeDebug_(String("Final RT tolerance: ") + rt_tol, 1);
//general statistics
std::vector<double> ints_t;
std::vector<double> ints_i;
std::vector<double> ints_found;
std::vector<double> ints_missed;
Map<String, UInt> abort_strings;
for (Size m = 0; m < features_truth.size(); ++m)
{
Feature& f_t = features_truth[m];
UInt match_count = 0;
bool correct_charge = false;
bool exact_centroid_match = false;
Size last_match_index = features_in.size() + 1;
for (Size a = 0; a < features_in.size(); ++a)
{
const Feature& f_i = features_in[a];
//RT match
if (fabs(f_i.getRT() - f_t.getRT()) < rt_tol)
{
double charge_mz_tol = mz_tol / f_t.getCharge();
//Exact m/z match
if (fabs(f_i.getMZ() - f_t.getMZ()) < charge_mz_tol)
{
++match_count;
exact_centroid_match = true;
if (f_i.getCharge() == f_t.getCharge()) correct_charge = true;
last_match_index = a;
}
//Centroid is one trace off, but still contained in the convex hull
else if (f_i.getConvexHull().getBoundingBox().encloses(f_t.getPosition())
&&
(
fabs(f_i.getMZ() + 1.0 / f_t.getCharge() - f_t.getMZ()) < charge_mz_tol
||
fabs(f_i.getMZ() - 1.0 / f_t.getCharge() - f_t.getMZ()) < charge_mz_tol
)
)
{
++match_count;
last_match_index = a;
if (f_i.getCharge() == f_t.getCharge()) correct_charge = true;
}
}
}
f_t.setMetaValue("matches", match_count);
if (match_count == 1)
{
//flag matched feature with addition information
if (correct_charge)
{
f_t.setMetaValue("correct_charge", String("true"));
f_t.setMetaValue("intensity_ratio", features_in[last_match_index].getIntensity() / f_t.getIntensity());
features_in[last_match_index].setMetaValue("correct_hit", "true"); //flag the feature for ROC curve
}
else
{
f_t.setMetaValue("correct_charge", String("false"));
}
if (exact_centroid_match)
{
f_t.setMetaValue("exact_centroid_match", String("true"));
}
else
{
f_t.setMetaValue("exact_centroid_match", String("false"));
}
}
//evaluation of correct features only
if (match_count == 1 && correct_charge)
{
ints_t.push_back(f_t.getIntensity());
ints_i.push_back(features_in[last_match_index].getIntensity());
ints_found.push_back(f_t.getIntensity());
}
else
{
ints_missed.push_back(f_t.getIntensity());
//look up the abort reason of the nearest seed
double best_score_ab = 0;
String reason = "";
for (Size b = 0; b < abort_reasons.size(); ++b)
{
const Feature& f_ab = abort_reasons[b];
if (fabs(f_ab.getRT() - f_t.getRT()) <= rt_tol
&& fabs(f_ab.getMZ() - f_t.getMZ()) <= mz_tol)
{
double score = (1.0 - fabs(f_ab.getMZ() - f_t.getMZ()) / mz_tol) * (1.0 - fabs(f_ab.getRT() - f_t.getRT()) / rt_tol);
if (score > best_score_ab)
{
best_score_ab = score;
reason = f_ab.getMetaValue("abort_reason");
}
}
}
if (reason == "")
{
reason = "No seed found";
}
if (abort_strings.has(reason))
{
abort_strings[reason]++;
}
else
{
abort_strings[reason] = 1;
}
}
}
//------------------------ general statistics ------------------------
cout << endl;
cout << "general information:" << endl;
cout << "====================" << endl;
cout << "input features: " << features_in.size() << endl;
cout << "truth features: " << features_truth.size() << endl;
//------------------------ matches ------------------------
cout << endl;
cout << "feature matching statistics:" << endl;
cout << "============================" << endl;
Size no_match = count(features_truth, "matches", "0");
cout << "no match: " << no_match << percentage(no_match, features_truth.size()) << endl;
Size one_match = count(features_truth, "matches", "1");
cout << "one match: " << one_match << percentage(one_match, features_truth.size()) << endl;
Size charge_match = count(features_truth, "correct_charge", "true");
cout << " - correct charge: " << charge_match << percentage(charge_match, features_truth.size()) << endl;
Size centroid_match = count(features_truth, "exact_centroid_match", "true");
cout << " - exact centroid match: " << centroid_match << percentage(centroid_match, features_truth.size()) << endl;
Size multi_match = features_truth.size() - count(features_truth, "matches", "0") - count(features_truth, "matches", "1");
cout << "multiple matches: " << multi_match << percentage(multi_match, features_truth.size()) << endl;
Size incorrect_match = multi_match + one_match - charge_match;
cout << "incorrect matches: " << incorrect_match << percentage(incorrect_match, features_truth.size()) << endl;
if (abort_reasons.size())
{
cout << "reasons for unmatched features:" << endl;
for (Map<String, UInt>::iterator it = abort_strings.begin(); it != abort_strings.end(); ++it)
{
cout << " - " << String(it->second).fillLeft(' ', 4) << ": " << it->first << endl;
}
}
//------------------------ intensity ------------------------
cout << endl;
cout << "intensity statistics:" << endl;
cout << "=====================" << endl;
if (ints_i.empty())
{
cout << "correlation of found features: nan" << endl;
}
else
{
cout << "correlation of found features: " << pearsonCorrelationCoefficient(ints_i.begin(), ints_i.end(), ints_t.begin(), ints_t.end()) << endl;
}
if (ints_found.empty())
{
cout << "intensity distribution of found: 0.0 0.0 0.0 0.0 0.0" << endl;
}
else
{
cout << "intensity distribution of found: " << fiveNumbers(ints_found, 1) << endl;
}
if (ints_missed.empty())
{
cout << "intensity distribution of missed: 0.0 0.0 0.0 0.0 0.0" << endl;
}
else
{
cout << "intensity distribution of missed: " << fiveNumbers(ints_missed, 1) << endl;
}
//------------------------ charges ------------------------
cout << endl;
cout << "charge matches statistics:" << endl;
cout << "===========================" << endl;
Map<UInt, UInt> present_charges, found_charges;
for (Size i = 0; i < features_truth.size(); ++i)
{
UInt charge = features_truth[i].getCharge();
present_charges[charge]++;
if (features_truth[i].getMetaValue("correct_charge").toString() == "true")
{
found_charges[charge]++;
}
}
for (Map<UInt, UInt>::const_iterator it = present_charges.begin(); it != present_charges.end(); ++it)
{
cout << "charge " << it->first << ": " << found_charges[it->first] << "/" << it->second << percentage(found_charges[it->first], it->second) << endl;
}
//write output
if (getStringOption_("out") != "")
{
FeatureXMLFile().store(getStringOption_("out"), features_truth);
}
//ROC curve
if (getStringOption_("out_roc") != "")
{
TextFile tf;
tf.addLine("false\tcorrect\tFDR\tTPR");
features_in.sortByIntensity(true);
UInt f_correct = 0;
UInt f_false = 0;
double found = features_in.size();
double correct = features_truth.size();
for (Size i = 0; i < features_in.size(); ++i)
{
if (features_in[i].metaValueExists("correct_hit"))
{
++f_correct;
}
else
{
++f_false;
}
tf.addLine(String(f_false) + "\t" + f_correct + "\t" + String::number(f_false / found, 3) + "\t" + String::number(f_correct / correct, 3));
}
tf.store(getStringOption_("out_roc"));
}
return EXECUTION_OK;
}
ExitCodes main_(int, const char**)
{
//----------------------------------------------------------------
// load data
//----------------------------------------------------------------
StringList in_list = getStringList_("in");
String out = getStringOption_("out");
String out_csv = getStringOption_("out_csv");
String format = getStringOption_("out_type");
if (out.empty() && out_csv.empty())
{
LOG_ERROR << "Neither 'out' nor 'out_csv' were provided. Please assign at least one of them." << std::endl;
return ILLEGAL_PARAMETERS;
}
if (!out.empty() && format == "") // get from filename
{
try
{
format = out.suffix('.');
}
catch (Exception::ElementNotFound& /*e*/)
{
format = "nosuffix";
}
// check if format is valid:
if (!ListUtils::contains(out_formats_, format.toLower()))
{
LOG_ERROR << "No explicit image output format was provided via 'out_type', and the suffix ('" << format << "') does not resemble a valid type. Please fix one of them." << std::endl;
return ILLEGAL_PARAMETERS;
}
}
double q_min = getDoubleOption_("q_min");
double q_max = getDoubleOption_("q_max");
if (q_min >= q_max)
{
LOG_ERROR << "The parameter 'q_min' must be smaller than 'q_max'. Quitting..." << std::endl;
return ILLEGAL_PARAMETERS;
}
IDEvaluationBase* mw = new IDEvaluationBase();
Param alg_param = mw->getParameters();
alg_param.insert("", getParam_().copy("algorithm:", true));
mw->setParameters(alg_param);
if (!mw->loadFiles(in_list))
{
LOG_ERROR << "Tool failed. See above." << std::endl;
return INCOMPATIBLE_INPUT_DATA;
}
mw->setVisibleArea(q_min, q_max);
if (!out.empty()) // save as image and exit
{
String error;
bool r = mw->exportAsImage(out.toQString(), error, format.toQString());
if (r) return EXECUTION_OK;
else
{
LOG_ERROR << error << std::endl;
return ILLEGAL_PARAMETERS;
}
}
if (!out_csv.empty())
{
TextFile tf;
for (Size i = 0; i < mw->getPoints().size(); ++i)
{
MSSpectrum s = mw->getPoints()[i];
StringList sl1;
StringList sl2;
for (Size j = 0; j < s.size(); ++j)
{
sl1.push_back(s[j].getMZ());
sl2.push_back(s[j].getIntensity());
}
tf.addLine(String("# ") + String(s.getMetaValue("search_engine")));
tf.addLine(ListUtils::concatenate(sl1, ","));
tf.addLine(ListUtils::concatenate(sl2, ","));
}
tf.store(out_csv);
}
delete(mw);
return EXECUTION_OK;
}
void IBSpectraFile::store(const String& filename, const ConsensusMap& cm)
{
// typdefs for shorter code
typedef std::vector<ProteinHit>::iterator ProtHitIt;
// general settings .. do we need to expose these?
// ----------------------------------------------------------------------
/// Allow also non-unique peptides to be exported
bool allow_non_unique = true;
/// Intensities below this value will be set to 0.0 to avoid numerical problems when quantifying
double intensity_threshold = 0.00001;
// ----------------------------------------------------------------------
// guess experiment type
boost::shared_ptr<IsobaricQuantitationMethod> quantMethod = guessExperimentType_(cm);
// we need the protein identifications to reference the protein names
ProteinIdentification protIdent;
bool has_proteinIdentifications = false;
if (cm.getProteinIdentifications().size() > 0)
{
protIdent = cm.getProteinIdentifications()[0];
has_proteinIdentifications = true;
}
// start the file by adding the tsv header
TextFile textFile;
textFile.addLine(ListUtils::concatenate(constructHeader_(*quantMethod), "\t"));
for (ConsensusMap::ConstIterator cm_iter = cm.begin();
cm_iter != cm.end();
++cm_iter)
{
const ConsensusFeature& cFeature = *cm_iter;
std::vector<IdCSV> entries;
/// 1st we extract the identification information from the consensus feature
if (cFeature.getPeptideIdentifications().size() == 0 || !has_proteinIdentifications)
{
// we store unidentified hits anyway, because the iTRAQ quant is still helpful for normalization
entries.push_back(IdCSV());
}
else
{
// protein name:
const PeptideHit& peptide_hit = cFeature.getPeptideIdentifications()[0].getHits()[0];
std::set<String> protein_accessions = peptide_hit.extractProteinAccessions();
if (protein_accessions.size() != 1)
{
if (!allow_non_unique) continue; // we only want unique peptides
}
for (std::set<String>::const_iterator prot_ac = protein_accessions.begin(); prot_ac != protein_accessions.end(); ++prot_ac)
{
IdCSV entry;
entry.charge = cFeature.getPeptideIdentifications()[0].getHits()[0].getCharge();
entry.peptide = cFeature.getPeptideIdentifications()[0].getHits()[0].getSequence().toUnmodifiedString();
entry.theo_mass = cFeature.getPeptideIdentifications()[0].getHits()[0].getSequence().getMonoWeight(Residue::Full, cFeature.getPeptideIdentifications()[0].getHits()[0].getCharge());
// write modif
entry.modif = getModifString_(cFeature.getPeptideIdentifications()[0].getHits()[0].getSequence());
ProtHitIt proteinHit = protIdent.findHit(*prot_ac);
if (proteinHit == protIdent.getHits().end())
{
std::cerr << "Protein referenced in peptide not found...\n";
continue; // protein not found
}
entry.accession = proteinHit->getAccession();
entries.push_back(entry);
}
}
// 2nd we add the quantitative information of the channels
// .. skip features with 0 intensity
if (cFeature.getIntensity() == 0)
{
continue;
}
for (std::vector<IdCSV>::iterator entry = entries.begin();
entry != entries.end();
++entry)
{
// set parent intensity
entry->parent_intens = cFeature.getIntensity();
entry->retention_time = cFeature.getRT();
entry->spectrum = cFeature.getUniqueId();
entry->exp_mass = cFeature.getMZ();
// create output line
StringList currentLine;
// add entry to currentLine
entry->toStringList(currentLine);
// extract channel intensities and positions
std::map<Int, double> intensityMap;
ConsensusFeature::HandleSetType features = cFeature.getFeatures();
for (ConsensusFeature::HandleSetType::const_iterator fIt = features.begin();
fIt != features.end();
++fIt)
{
intensityMap[Int(fIt->getMZ())] = (fIt->getIntensity() > intensity_threshold ? fIt->getIntensity() : 0.0);
}
for (IsobaricQuantitationMethod::IsobaricChannelList::const_iterator it = quantMethod->getChannelInformation().begin();
it != quantMethod->getChannelInformation().end();
++it)
{
currentLine.push_back(String(it->center));
}
for (IsobaricQuantitationMethod::IsobaricChannelList::const_iterator it = quantMethod->getChannelInformation().begin();
it != quantMethod->getChannelInformation().end();
++it)
{
currentLine.push_back(String(intensityMap[int(it->center)]));
}
textFile.addLine(ListUtils::concatenate(currentLine, "\t"));
}
}
// write to file
textFile.store(filename);
}
ExitCodes main_(int, const char**)
{
StringList cv_files = getStringList_("cv_files");
StringList cv_names = getStringList_("cv_names");
if (cv_files.size() != cv_names.size())
{
cerr << "Error: You have to specify an identifier for each CV file. Aborting!" << endl;
return ILLEGAL_PARAMETERS;
}
// load cv terms
ControlledVocabulary cv;
for (Size i = 0; i < cv_files.size(); ++i)
{
cv.loadFromOBO(cv_names[i], cv_files[i]);
}
Map<String, ControlledVocabulary::CVTerm> terms = cv.getTerms();
// load mappings from mapping file
String mapping_file = getStringOption_("mapping_file");
CVMappings mappings;
CVMappingFile().load(mapping_file, mappings);
//store HTML version of mapping and CV
if (getStringOption_("html") != "")
{
TextFile file;
file.addLine("<HTML>");
file.addLine(" <HEAD>");
file.addLine(" <TITLE>CV mapping file</TITLE>");
file.addLine(" <SCRIPT language=javascript type='text/javascript'>");
file.addLine(" function toggleDiv(layer_ref,force_state) ");
file.addLine(" {");
file.addLine(" if (document.getElementById(layer_ref).style.display=='none' || force_state=='true')");
file.addLine(" {");
file.addLine(" document.getElementById(layer_ref).style.display = 'block';");
file.addLine(" }");
file.addLine(" else if (document.getElementById(layer_ref).style.display=='block' || force_state=='false')");
file.addLine(" {");
file.addLine(" document.getElementById(layer_ref).style.display = 'none';");
file.addLine(" }");
file.addLine(" }");
file.addLine(" </SCRIPT>");
file.addLine(" </HEAD>");
file.addLine(" <BODY>");
//count the number of terms and add button to expend/collaps all terms
Int term_count = 0;
for (vector<CVMappingRule>::const_iterator it = mappings.getMappingRules().begin(); it != mappings.getMappingRules().end(); ++it)
{
for (vector<CVMappingTerm>::const_iterator tit = it->getCVTerms().begin(); tit != it->getCVTerms().end(); ++tit)
{
++term_count;
}
}
String expand_all = " <a href=\"javascript:toggleDiv('div0','true')";
String collapse_all = " <a href=\"javascript:toggleDiv('div0','false')";
for (Int i = 1; i < term_count; ++i)
{
expand_all += String(";toggleDiv('div") + i + "','true')";
collapse_all += String(";toggleDiv('div") + i + "','false')";
}
file.addLine(expand_all + "\">Expand all</a><BR>");
file.addLine(collapse_all + "\">Collapse all</a>");
file.addLine(" <TABLE width=100% border=0>");
term_count = -1;
for (vector<CVMappingRule>::const_iterator it = mappings.getMappingRules().begin(); it != mappings.getMappingRules().end(); ++it)
{
//create rule line
file.addLine(" <TR><TD colspan=\"2\"><HR></TD></TR>");
file.addLine(String(" <TR><TD>Identifier:</TD><TD><B>") + it->getIdentifier() + "</B></TD></TR>");
file.addLine(String(" <TR><TD>Element:</TD><TD><B>") + it->getElementPath() + "</B></TD></TR>");
if (it->getRequirementLevel() == CVMappingRule::MUST)
{
file.addLine(" <TR><TD>Requirement level:</TD><TD><FONT color=\"red\">MUST</FONT></TD></TR>");
}
else if (it->getRequirementLevel() == CVMappingRule::SHOULD)
{
file.addLine(" <TR><TD>Requirement level:</TD><TD><FONT color=\"orange\">SHOULD</FONT></TD></TR>");
}
else if (it->getRequirementLevel() == CVMappingRule::MAY)
{
file.addLine(" <TR><TD>Requirement level:</TD><TD><FONT color=\"green\">MAY</FONT></TD></TR>");
}
if (it->getCombinationsLogic() == CVMappingRule::AND)
{
file.addLine(" <TR><TD>Combination logic:</TD><TD><FONT color=\"red\">AND</FONT></TD></TR>");
}
else if (it->getCombinationsLogic() == CVMappingRule::XOR)
{
file.addLine(" <TR><TD>Combination logic:</TD><TD><FONT color=\"orange\">XOR</FONT></TD></TR>");
}
else if (it->getCombinationsLogic() == CVMappingRule::OR)
{
file.addLine(" <TR><TD>Combination logic:</TD><TD><FONT color=\"green\">OR</FONT></TD></TR>");
}
//create table with terms
for (vector<CVMappingTerm>::const_iterator tit = it->getCVTerms().begin(); tit != it->getCVTerms().end(); ++tit)
{
//create term line
String term_line = String(" <TR><TD valign=\"top\">Term:</TD><TD>");
if (tit->getAllowChildren())
{
++term_count;
term_line += String("<a href=\"javascript:toggleDiv('div") + term_count + "','')\" style=\"text-decoration:none\" >+</a> ";
}
else
{
term_line += String(" ");
}
//add Term accession, name and description (as popup)
if (cv.exists(tit->getAccession()))
{
const ControlledVocabulary::CVTerm& child_term = cv.getTerm(tit->getAccession());
String description = child_term.description;
if (child_term.synonyms.size() != 0)
{
description += String(" -- Synonyms: '") + ListUtils::concatenate(child_term.synonyms, ", ") + "'";
}
term_line += "<span title=\"" + description + "\">";
}
term_line += tit->getAccession() + " ! " + tit->getTermName();
if (cv.exists(tit->getAccession()))
{
term_line += "</span>";
//check if term accession and term name correspond to the CV
const ControlledVocabulary::CVTerm& main_term = cv.getTerm(tit->getAccession());
if (main_term.name != tit->getTermName())
{
cerr << "Warning: Accession '" << tit->getAccession() << "' and name '" << tit->getTermName() << "' do not match. Name should be '" << main_term.name << "'." << endl;
}
}
//tags
StringList tags;
if (!tit->getUseTerm())
{
tags.push_back("children only");
}
if (tit->getIsRepeatable())
{
tags.push_back("repeatable");
}
if (cv.exists(tit->getAccession()))
{
const ControlledVocabulary::CVTerm& term = cv.getTerm(tit->getAccession());
if (term.obsolete)
{
tags.push_back("<font color=darkred>obsolete</font>");
}
if (term.xref_type != ControlledVocabulary::CVTerm::NONE)
{
tags.push_back("value-type=" + ControlledVocabulary::CVTerm::getXRefTypeName(term.xref_type));
}
if (term.units.size() > 0)
{
StringList units;
for (set<String>::const_iterator u_it = term.units.begin(); u_it != term.units.end(); ++u_it)
{
units.push_back(*u_it + "!" + cv.getTerm(*u_it).name);
}
tags.push_back(String("units=") + ListUtils::concatenate(units, ","));
}
if (term.xref_binary.size() > 0)
{
StringList types;
for (StringList::const_iterator u_it = term.xref_binary.begin(); u_it != term.xref_binary.end(); ++u_it)
{
types.push_back(*u_it + "!" + cv.getTerm(*u_it).name);
}
tags.push_back(String("binary-array-types=") + ListUtils::concatenate(types, ","));
}
}
if (tags.size() != 0)
{
term_line += String("<FONT color=\"grey\"> (") + ListUtils::concatenate(tags, ", ") + ")</FONT>";
}
file.addLine(term_line);
// check whether we need the whole tree, or just the term itself
if (tit->getAllowChildren())
{
file.addLine(String(" <div id=\"div") + term_count + "\" style=\"display: none\">");
if (cv.exists(tit->getAccession()))
{
writeTermTree_(tit->getAccession(), cv, file, 1);
//BEGIN - THIS IS NEEDED FOR WRITING PARSERS ONLY
/*
set<String> allowed_terms;
cv.getAllChildTerms(allowed_terms, tit->getAccession());
for (set<String>::const_iterator atit=allowed_terms.begin(); atit!=allowed_terms.end(); ++atit)
{
const ControlledVocabulary::CVTerm& child_term = cv.getTerm(*atit);
String parser_string = String("os << \"<cvParam cvRef=\\\"MS\\\" accession=\\\"") + child_term.id + "\\\" name=\\\"" + child_term.name + "\\\"";
for (Size i=0; i<child_term.unparsed.size(); ++i)
{
//TODO this does not work anymore. The type is now stored as a member
if (child_term.unparsed[i].hasSubstring("value-type:xsd\\:int") || child_term.unparsed[i].hasSubstring("value-type:xsd\\:float") || child_term.unparsed[i].hasSubstring("value-type:xsd\\:string"))
{
parser_string += " value=\\\"\" << << \"\\\"";
}
}
parser_string += "/>\\n\";<BR>";
file.push_back(parser_string);
}*/
}
else
{
file.addLine(" - Missing terms, CV not loaded...");
cerr << "Warning: no child terms for " << tit->getAccession() << " found!" << endl;
}
file.addLine(" </div>");
file.addLine(" </TD></TD></TR>");
}
}
}
file.addLine(" </TABLE>");
file.addLine(" </BODY>");
file.addLine("</HTML>");
file.store(getStringOption_("html"));
return EXECUTION_OK;
}
// iterator over all mapping rules and store the mentioned terms
StringList ignore_namespaces = getStringList_("ignore_cv");
set<String> ignore_cv_list;
for (StringList::const_iterator it = ignore_namespaces.begin(); it != ignore_namespaces.end(); ++it)
{
ignore_cv_list.insert(*it);
}
set<String> used_terms;
for (vector<CVMappingRule>::const_iterator it = mappings.getMappingRules().begin(); it != mappings.getMappingRules().end(); ++it)
{
set<String> allowed_terms;
// iterate over all allowed terms
for (vector<CVMappingTerm>::const_iterator tit = it->getCVTerms().begin(); tit != it->getCVTerms().end(); ++tit)
{
// check whether the term itself it allowed, or only its children
if (tit->getUseTerm())
{
allowed_terms.insert(tit->getAccession());
}
// check whether we need the whole tree, or just the term itself
if (tit->getAllowChildren())
{
// check whether we want to ignore this term
if (!(tit->getAccession().has(':') && ignore_cv_list.find(tit->getAccession().prefix(':')) != ignore_cv_list.end()))
{
cv.getAllChildTerms(allowed_terms, tit->getAccession());
}
// also add the term itself to the used_terms, because all the children are allowed
used_terms.insert(tit->getAccession());
}
}
// print the allowed terms for the rule
cout << "MappingRule: id=" << it->getIdentifier() << ", elementPath=" << it->getElementPath() << ", #terms=" << it->getCVTerms().size() << endl;
for (set<String>::const_iterator ait = allowed_terms.begin(); ait != allowed_terms.end(); ++ait)
{
cout << *ait << " " << terms[*ait].name << endl;
}
used_terms.insert(allowed_terms.begin(), allowed_terms.end());
}
// find unused terms, which CANNOT be used in the XML due to the mapping file
set<String> unused_terms;
for (Map<String, ControlledVocabulary::CVTerm>::ConstIterator it = terms.begin(); it != terms.end(); ++it)
{
if (used_terms.find(it->first) == used_terms.end())
{
unused_terms.insert(it->first);
}
}
cout << "\n\nCVTerms which are unused in the mapping file and therefore MUST NOT be used in an instance document" << endl;
for (set<String>::const_iterator it = unused_terms.begin(); it != unused_terms.end(); ++it)
{
cout << *it << " " << terms[*it].name;
// print also parent names
for (set<String>::const_iterator pit = terms[*it].parents.begin(); pit != terms[*it].parents.end(); ++pit)
{
cout << " " << terms[*pit].id << " " << terms[*pit].name;
}
cout << endl;
}
return EXECUTION_OK;
}
