void IDFilter::filterIdentificationsByExclusionPeptides(const PeptideIdentification& identification,
const set<String>& peptides,
PeptideIdentification& filtered_identification)
{
String protein_sequences;
String accession_sequences;
vector<PeptideHit> filtered_peptide_hits;
PeptideHit temp_peptide_hit;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
for (Size i = 0; i < identification.getHits().size(); i++)
{
if (find(peptides.begin(), peptides.end(), identification.getHits()[i].getSequence().toString()) == peptides.end())
{
filtered_peptide_hits.push_back(identification.getHits()[i]);
}
}
if (!filtered_peptide_hits.empty())
{
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
}
void IDFilter::filterIdentificationsByCharge(const PeptideIdentification& identification,
Int min_charge,
PeptideIdentification& filtered_identification)
{
vector<Size> new_peptide_indices;
vector<PeptideHit> filtered_peptide_hits;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
const vector<PeptideHit>& temp_peptide_hits = identification.getHits();
for (Size i = 0; i < temp_peptide_hits.size(); i++)
{
if (temp_peptide_hits[i].getCharge() >= min_charge)
{
new_peptide_indices.push_back(i);
}
}
for (Size i = 0; i < new_peptide_indices.size(); i++)
{
filtered_peptide_hits.push_back(identification.getHits()[new_peptide_indices[i]]);
}
if (!filtered_peptide_hits.empty())
{
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
}
void IDFilter::filterIdentificationsByRTPValues(const PeptideIdentification& identification,
PeptideIdentification& filtered_identification,
DoubleReal p_value)
{
DoubleReal border = 1 - p_value;
vector<PeptideHit> filtered_peptide_hits;
PeptideHit temp_peptide_hit;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
Size missing_meta_value = 0;
for (Size i = 0; i < identification.getHits().size(); i++)
{
if (identification.getHits()[i].metaValueExists("predicted_RT_p_value"))
{
if ((DoubleReal)(identification.getHits()[i].getMetaValue("predicted_RT_p_value")) <= border)
{
filtered_peptide_hits.push_back(identification.getHits()[i]);
}
}
else
++missing_meta_value;
}
if (missing_meta_value > 0)
LOG_WARN << "Filtering identifications by p-value did not work on " << missing_meta_value << " of " << identification.getHits().size() << " hits. Your data is missing a meta-value ('predicted_RT_p_value') from RTPredict!\n";
if (!filtered_peptide_hits.empty())
{
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
}
void IDFilter::filterIdentificationsByProteins(const PeptideIdentification& identification,
const vector<FASTAFile::FASTAEntry>& proteins,
PeptideIdentification& filtered_identification,
bool no_protein_identifiers)
{
// TODO: this is highly inefficient! the Protein-Index should be build once for all peptide-identifications instead of
// doing this once for every ID. Furthermore the index itself is inefficient (use seqan instead)
String protein_sequences;
String accession_sequences;
vector<PeptideHit> filtered_peptide_hits;
PeptideHit temp_peptide_hit;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
for (Size i = 0; i < proteins.size(); i++)
{
if (proteins[i].identifier != "")
{
accession_sequences.append("*" + proteins[i].identifier);
}
if (proteins[i].sequence != "")
{
protein_sequences.append("*" + proteins[i].sequence);
}
}
accession_sequences.append("*");
protein_sequences.append("*");
for (Size i = 0; i < identification.getHits().size(); i++)
{
if (no_protein_identifiers || accession_sequences == "*") // filter by sequence alone if no protein accesssions are available
{
if (protein_sequences.find(identification.getHits()[i].getSequence().toUnmodifiedString()) != String::npos)
{
filtered_peptide_hits.push_back(identification.getHits()[i]);
}
}
else // filter by protein accessions
{
for (vector<String>::const_iterator ac_it = identification.getHits()[i].getProteinAccessions().begin();
ac_it != identification.getHits()[i].getProteinAccessions().end();
++ac_it)
{
if (accession_sequences.find("*" + *ac_it) != String::npos)
{
filtered_peptide_hits.push_back(identification.getHits()[i]);
break; // we found a matching protein, the peptide is valid -> exit
}
}
}
}
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
void IDFilter::filterIdentificationsByBestHits(const PeptideIdentification& identification,
PeptideIdentification& filtered_identification,
bool strict)
{
vector<PeptideHit> filtered_peptide_hits;
PeptideHit temp_peptide_hit;
vector<Size> new_peptide_indices;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
if (!identification.getHits().empty())
{
Real optimal_value = identification.getHits()[0].getScore();
new_peptide_indices.push_back(0);
// searching for peptide(s) with maximal score
for (Size i = 1; i < identification.getHits().size(); i++)
{
Real temp_score = identification.getHits()[i].getScore();
bool new_leader = false;
if ((identification.isHigherScoreBetter() && (temp_score > optimal_value))
|| (!identification.isHigherScoreBetter() && (temp_score < optimal_value)))
new_leader = true;
if (new_leader)
{
optimal_value = temp_score;
new_peptide_indices.clear();
new_peptide_indices.push_back(i);
}
else if (temp_score == optimal_value)
{
new_peptide_indices.push_back(i);
}
}
if (!strict || new_peptide_indices.size() == 1)
{
for (Size i = 0; i < new_peptide_indices.size(); i++)
{
filtered_peptide_hits.push_back(identification.getHits()[new_peptide_indices[i]]);
}
}
}
if (!filtered_peptide_hits.empty())
{
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
}
void IDFilter::filterIdentificationsByMzError(const PeptideIdentification& identification, DoubleReal mass_error, bool unit_ppm, PeptideIdentification& filtered_identification)
{
vector<PeptideHit> hits;
filtered_identification = identification;
vector<PeptideHit> temp_hits = identification.getHits();
for (vector<PeptideHit>::iterator it = temp_hits.begin(); it != temp_hits.end(); ++it)
{
Int charge = it->getCharge();
if (charge == 0)
{
charge = 1;
}
DoubleReal exp_mz = (DoubleReal)identification.getMetaValue("MZ");
DoubleReal theo_mz = (it->getSequence().getMonoWeight() + (DoubleReal)charge * Constants::PROTON_MASS_U) / (DoubleReal)charge;
DoubleReal error(exp_mz - theo_mz);
if (unit_ppm)
{
error = error / theo_mz * (DoubleReal)1e6;
}
if (fabs(error) <= mass_error)
{
hits.push_back(*it);
}
}
filtered_identification.setHits(hits);
}
void IDFilter::filterIdentificationsByExclusionPeptides(const PeptideIdentification& identification,
const set<String>& peptides,
bool ignore_modifications,
PeptideIdentification& filtered_identification)
{
vector<PeptideHit> filtered_peptide_hits;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
for (Size i = 0; i < identification.getHits().size(); i++)
{
String query = ignore_modifications ? identification.getHits()[i].getSequence().toUnmodifiedString() : identification.getHits()[i].getSequence().toString();
if (find(peptides.begin(), peptides.end(), query) == peptides.end())
{
filtered_peptide_hits.push_back(identification.getHits()[i]);
}
}
if (!filtered_peptide_hits.empty())
{
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
}
std::vector<PeptideIdentification> toPepVec(const QStringList& sl_pep)
{
std::vector<PeptideIdentification> pep_vec;
for (Size i = 0; i < sl_pep.size(); ++i)
{
PeptideHit hit;
hit.setSequence(AASequence::fromString(sl_pep[int(i)]));
std::vector<PeptideHit> hits;
hits.push_back(hit);
PeptideIdentification pi;
pi.setHits(hits);
pep_vec.push_back(pi);
}
return pep_vec;
}
void IDFilter::filterIdentificationsByLength(const PeptideIdentification& identification,
PeptideIdentification& filtered_identification,
Size min_length,
Size max_length)
{
vector<Size> new_peptide_indices;
vector<PeptideHit> filtered_peptide_hits;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
Size ml = max_length;
if (max_length < min_length)
{
ml = UINT_MAX;
}
const vector<PeptideHit>& temp_peptide_hits = identification.getHits();
for (Size i = 0; i < temp_peptide_hits.size(); i++)
{
if (temp_peptide_hits[i].getSequence().size() >= min_length && temp_peptide_hits[i].getSequence().size() <= ml)
{
new_peptide_indices.push_back(i);
}
}
for (Size i = 0; i < new_peptide_indices.size(); i++)
{
filtered_peptide_hits.push_back(identification.getHits()[new_peptide_indices[i]]);
}
if (!filtered_peptide_hits.empty())
{
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
}
void IDFilter::filterIdentificationsUnique(const PeptideIdentification& identification,
PeptideIdentification& filtered_identification)
{
vector<PeptideHit> hits;
filtered_identification = identification;
vector<PeptideHit> temp_hits = identification.getHits();
for (vector<PeptideHit>::iterator it = temp_hits.begin();
it != temp_hits.end();
++it)
{
if (find(hits.begin(), hits.end(), *it) == hits.end())
{
hits.push_back(*it);
}
}
filtered_identification.setHits(hits);
}
void IDFilter::filterIdentificationsByCharge(const PeptideIdentification& identification,
Int min_charge,
PeptideIdentification& filtered_identification)
{
filtered_identification = identification;
const vector<PeptideHit>& temp_peptide_hits = identification.getHits();
vector<PeptideHit> filtered_peptide_hits;
for (Size i = 0; i < temp_peptide_hits.size(); ++i)
{
if (temp_peptide_hits[i].getCharge() >= min_charge)
{
filtered_peptide_hits.push_back(temp_peptide_hits[i]);
}
}
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
void IDFilter::filterIdentificationsUnique(const PeptideIdentification& identification,
PeptideIdentification& filtered_identification)
{
// there's no "PeptideHit::operator<" defined, so we can't use a set nor
// "sort" + "unique" from the standard library
vector<PeptideHit> hits;
filtered_identification = identification;
vector<PeptideHit> temp_hits = identification.getHits();
for (vector<PeptideHit>::iterator it = temp_hits.begin();
it != temp_hits.end();
++it)
{
if (find(hits.begin(), hits.end(), *it) == hits.end())
{
hits.push_back(*it);
}
}
filtered_identification.setHits(hits);
}
void IDFilter::filterIdentificationsByLength(const PeptideIdentification& identification,
PeptideIdentification& filtered_identification,
Size min_length,
Size max_length)
{
filtered_identification = identification;
if (max_length < min_length)
{
max_length = UINT_MAX;
}
const vector<PeptideHit>& temp_peptide_hits = identification.getHits();
vector<PeptideHit> filtered_peptide_hits;
for (Size i = 0; i < temp_peptide_hits.size(); ++i)
{
if (min_length <= temp_peptide_hits[i].getSequence().size() && temp_peptide_hits[i].getSequence().size() <= max_length)
{
filtered_peptide_hits.push_back(temp_peptide_hits[i]);
}
}
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
void IDFilter::filterIdentificationsByVariableModifications(const PeptideIdentification& identification,
const vector<String>& fixed_modifications,
PeptideIdentification& filtered_identification)
{
vector<Size> new_peptide_indices;
vector<PeptideHit> filtered_peptide_hits;
filtered_identification = identification;
filtered_identification.setHits(vector<PeptideHit>());
const vector<PeptideHit>& temp_peptide_hits = identification.getHits();
for (Size i = 0; i < temp_peptide_hits.size(); i++)
{
const AASequence& aa_seq = temp_peptide_hits[i].getSequence();
/*
TODO: check these cases
// check terminal modifications
if (aa_seq.hasNTerminalModification())
{
String unimod_name = aa_seq.getNTerminalModification();
if (find(fixed_modifications.begin(), fixed_modifications.end(), unimod_name) == fixed_modifications.end())
{
new_peptide_indices.push_back(i);
continue;
}
}
if (aa_seq.hasCTerminalModification())
{
String unimod_name = aa_seq.getCTerminalModification();
if (find(fixed_modifications.begin(), fixed_modifications.end(), unimod_name) == fixed_modifications.end())
{
new_peptide_indices.push_back(i);
continue;
}
}
*/
// check internal modifications
for (Size j = 0; j != aa_seq.size(); ++j)
{
if (aa_seq[j].isModified())
{
String unimod_name = aa_seq[j].getModification() + " (" + aa_seq[j].getOneLetterCode() + ")";
if (find(fixed_modifications.begin(), fixed_modifications.end(), unimod_name) == fixed_modifications.end())
{
new_peptide_indices.push_back(i);
continue;
}
}
}
}
for (Size i = 0; i < new_peptide_indices.size(); i++)
{
const PeptideHit& ph = temp_peptide_hits[new_peptide_indices[i]];
filtered_peptide_hits.push_back(ph);
}
if (!filtered_peptide_hits.empty())
{
filtered_identification.setHits(filtered_peptide_hits);
filtered_identification.assignRanks();
}
}
ExitCodes main_(int, const char**)
{
//-------------------------------------------------------------
// parameter handling
//-------------------------------------------------------------
StringList in_spec = getStringList_("in");
StringList out = getStringList_("out");
String in_lib = getStringOption_("lib");
String compare_function = getStringOption_("compare_function");
Int precursor_mass_multiplier = getIntOption_("round_precursor_to_integer");
float precursor_mass_tolerance = getDoubleOption_("precursor_mass_tolerance");
//Int min_precursor_charge = getIntOption_("min_precursor_charge");
//Int max_precursor_charge = getIntOption_("max_precursor_charge");
float remove_peaks_below_threshold = getDoubleOption_("filter:remove_peaks_below_threshold");
UInt min_peaks = getIntOption_("filter:min_peaks");
UInt max_peaks = getIntOption_("filter:max_peaks");
Int cut_peaks_below = getIntOption_("filter:cut_peaks_below");
StringList fixed_modifications = getStringList_("fixed_modifications");
StringList variable_modifications = getStringList_("variable_modifications");
Int top_hits = getIntOption_("top_hits");
if (top_hits < -1)
{
writeLog_("top_hits (should be >= -1 )");
return ILLEGAL_PARAMETERS;
}
//-------------------------------------------------------------
// loading input
//-------------------------------------------------------------
if (out.size() != in_spec.size())
{
writeLog_("out (should be as many as input files)");
return ILLEGAL_PARAMETERS;
}
time_t prog_time = time(NULL);
MSPFile spectral_library;
RichPeakMap query, library;
//spectrum which will be identified
MzMLFile spectra;
spectra.setLogType(log_type_);
time_t start_build_time = time(NULL);
//-------------------------------------------------------------
//building map for faster search
//-------------------------------------------------------------
//library containing already identified peptide spectra
vector<PeptideIdentification> ids;
spectral_library.load(in_lib, ids, library);
map<Size, vector<PeakSpectrum> > MSLibrary;
{
RichPeakMap::iterator s;
vector<PeptideIdentification>::iterator i;
ModificationsDB* mdb = ModificationsDB::getInstance();
for (s = library.begin(), i = ids.begin(); s < library.end(); ++s, ++i)
{
double precursor_MZ = (*s).getPrecursors()[0].getMZ();
Size MZ_multi = (Size)precursor_MZ * precursor_mass_multiplier;
map<Size, vector<PeakSpectrum> >::iterator found;
found = MSLibrary.find(MZ_multi);
PeakSpectrum librar;
bool variable_modifications_ok = true;
bool fixed_modifications_ok = true;
const AASequence& aaseq = i->getHits()[0].getSequence();
//variable fixed modifications
if (!fixed_modifications.empty())
{
for (Size i = 0; i < aaseq.size(); ++i)
{
const Residue& mod = aaseq.getResidue(i);
for (Size s = 0; s < fixed_modifications.size(); ++s)
{
if (mod.getOneLetterCode() == mdb->getModification(fixed_modifications[s]).getOrigin() && fixed_modifications[s] != mod.getModification())
{
fixed_modifications_ok = false;
break;
}
}
}
}
//variable modifications
if (aaseq.isModified() && (!variable_modifications.empty()))
{
for (Size i = 0; i < aaseq.size(); ++i)
{
if (aaseq.isModified(i))
{
const Residue& mod = aaseq.getResidue(i);
for (Size s = 0; s < variable_modifications.size(); ++s)
{
if (mod.getOneLetterCode() == mdb->getModification(variable_modifications[s]).getOrigin() && variable_modifications[s] != mod.getModification())
{
variable_modifications_ok = false;
break;
}
}
}
}
}
if (variable_modifications_ok && fixed_modifications_ok)
{
PeptideIdentification& translocate_pid = *i;
librar.getPeptideIdentifications().push_back(translocate_pid);
librar.setPrecursors(s->getPrecursors());
//library entry transformation
for (UInt l = 0; l < s->size(); ++l)
{
Peak1D peak;
if ((*s)[l].getIntensity() > remove_peaks_below_threshold)
{
const String& info = (*s)[l].getMetaValue("MSPPeakInfo");
if (info[0] == '?')
{
peak.setIntensity(sqrt(0.2 * (*s)[l].getIntensity()));
}
else
{
peak.setIntensity(sqrt((*s)[l].getIntensity()));
}
peak.setMZ((*s)[l].getMZ());
peak.setPosition((*s)[l].getPosition());
librar.push_back(peak);
}
}
if (found != MSLibrary.end())
{
found->second.push_back(librar);
}
else
{
vector<PeakSpectrum> tmp;
tmp.push_back(librar);
MSLibrary.insert(make_pair(MZ_multi, tmp));
}
}
}
}
time_t end_build_time = time(NULL);
cout << "Time needed for preprocessing data: " << (end_build_time - start_build_time) << "\n";
//compare function
PeakSpectrumCompareFunctor* comparor = Factory<PeakSpectrumCompareFunctor>::create(compare_function);
//-------------------------------------------------------------
// calculations
//-------------------------------------------------------------
double score;
StringList::iterator in, out_file;
for (in = in_spec.begin(), out_file = out.begin(); in < in_spec.end(); ++in, ++out_file)
{
time_t start_time = time(NULL);
spectra.load(*in, query);
//Will hold valuable hits
vector<PeptideIdentification> peptide_ids;
vector<ProteinIdentification> protein_ids;
// Write parameters to ProteinIdentifcation
ProteinIdentification prot_id;
//Parameters of identificaion
prot_id.setIdentifier("test");
prot_id.setSearchEngineVersion("SpecLibSearcher");
prot_id.setDateTime(DateTime::now());
prot_id.setScoreType(compare_function);
ProteinIdentification::SearchParameters searchparam;
searchparam.precursor_tolerance = precursor_mass_tolerance;
prot_id.setSearchParameters(searchparam);
/***********SEARCH**********/
for (UInt j = 0; j < query.size(); ++j)
{
//Set identifier for each identifications
PeptideIdentification pid;
pid.setIdentifier("test");
pid.setScoreType(compare_function);
ProteinHit pr_hit;
pr_hit.setAccession(j);
prot_id.insertHit(pr_hit);
//RichPeak1D to Peak1D transformation for the compare function query
PeakSpectrum quer;
bool peak_ok = true;
query[j].sortByIntensity(true);
double min_high_intensity = 0;
if (query[j].empty() || query[j].getMSLevel() != 2)
{
continue;
}
if (query[j].getPrecursors().empty())
{
writeLog_("Warning MS2 spectrum without precursor information");
continue;
}
min_high_intensity = (1 / cut_peaks_below) * query[j][0].getIntensity();
query[j].sortByPosition();
for (UInt k = 0; k < query[j].size() && k < max_peaks; ++k)
{
if (query[j][k].getIntensity() > remove_peaks_below_threshold && query[j][k].getIntensity() >= min_high_intensity)
{
Peak1D peak;
peak.setIntensity(sqrt(query[j][k].getIntensity()));
peak.setMZ(query[j][k].getMZ());
peak.setPosition(query[j][k].getPosition());
quer.push_back(peak);
}
}
if (quer.size() >= min_peaks)
{
peak_ok = true;
}
else
{
peak_ok = false;
}
double query_MZ = query[j].getPrecursors()[0].getMZ();
if (peak_ok)
{
bool charge_one = false;
Int percent = (Int) Math::round((query[j].size() / 100.0) * 3.0);
Int margin = (Int) Math::round((query[j].size() / 100.0) * 1.0);
for (vector<RichPeak1D>::iterator peak = query[j].end() - 1; percent >= 0; --peak, --percent)
{
if (peak->getMZ() < query_MZ)
{
break;
}
}
if (percent > margin)
{
charge_one = true;
}
float min_MZ = (query_MZ - precursor_mass_tolerance) * precursor_mass_multiplier;
float max_MZ = (query_MZ + precursor_mass_tolerance) * precursor_mass_multiplier;
for (Size mz = (Size)min_MZ; mz <= ((Size)max_MZ) + 1; ++mz)
{
map<Size, vector<PeakSpectrum> >::iterator found;
found = MSLibrary.find(mz);
if (found != MSLibrary.end())
{
vector<PeakSpectrum>& library = found->second;
for (Size i = 0; i < library.size(); ++i)
{
float this_MZ = library[i].getPrecursors()[0].getMZ() * precursor_mass_multiplier;
if (this_MZ >= min_MZ && max_MZ >= this_MZ && ((charge_one == true && library[i].getPeptideIdentifications()[0].getHits()[0].getCharge() == 1) || charge_one == false))
{
PeptideHit hit = library[i].getPeptideIdentifications()[0].getHits()[0];
PeakSpectrum& librar = library[i];
//Special treatment for SpectraST score as it computes a score based on the whole library
if (compare_function == "SpectraSTSimilarityScore")
{
SpectraSTSimilarityScore* sp = static_cast<SpectraSTSimilarityScore*>(comparor);
BinnedSpectrum quer_bin = sp->transform(quer);
BinnedSpectrum librar_bin = sp->transform(librar);
score = (*sp)(quer, librar); //(*sp)(quer_bin,librar_bin);
double dot_bias = sp->dot_bias(quer_bin, librar_bin, score);
hit.setMetaValue("DOTBIAS", dot_bias);
}
else
{
score = (*comparor)(quer, librar);
}
DataValue RT(library[i].getRT());
DataValue MZ(library[i].getPrecursors()[0].getMZ());
hit.setMetaValue("RT", RT);
hit.setMetaValue("MZ", MZ);
hit.setScore(score);
PeptideEvidence pe;
pe.setProteinAccession(pr_hit.getAccession());
hit.addPeptideEvidence(pe);
pid.insertHit(hit);
}
}
}
}
}
pid.setHigherScoreBetter(true);
pid.sort();
if (compare_function == "SpectraSTSimilarityScore")
{
if (!pid.empty() && !pid.getHits().empty())
{
vector<PeptideHit> final_hits;
final_hits.resize(pid.getHits().size());
SpectraSTSimilarityScore* sp = static_cast<SpectraSTSimilarityScore*>(comparor);
Size runner_up = 1;
for (; runner_up < pid.getHits().size(); ++runner_up)
{
if (pid.getHits()[0].getSequence().toUnmodifiedString() != pid.getHits()[runner_up].getSequence().toUnmodifiedString() || runner_up > 5)
{
break;
}
}
double delta_D = sp->delta_D(pid.getHits()[0].getScore(), pid.getHits()[runner_up].getScore());
for (Size s = 0; s < pid.getHits().size(); ++s)
{
final_hits[s] = pid.getHits()[s];
final_hits[s].setMetaValue("delta D", delta_D);
final_hits[s].setMetaValue("dot product", pid.getHits()[s].getScore());
final_hits[s].setScore(sp->compute_F(pid.getHits()[s].getScore(), delta_D, pid.getHits()[s].getMetaValue("DOTBIAS")));
//final_hits[s].removeMetaValue("DOTBIAS");
}
pid.setHits(final_hits);
pid.sort();
pid.setMZ(query[j].getPrecursors()[0].getMZ());
pid.setRT(query_MZ);
}
}
if (top_hits != -1 && (UInt)top_hits < pid.getHits().size())
{
vector<PeptideHit> hits;
hits.resize(top_hits);
for (Size i = 0; i < (UInt)top_hits; ++i)
{
hits[i] = pid.getHits()[i];
}
pid.setHits(hits);
}
peptide_ids.push_back(pid);
}
protein_ids.push_back(prot_id);
//-------------------------------------------------------------
// writing output
//-------------------------------------------------------------
IdXMLFile id_xml_file;
id_xml_file.store(*out_file, protein_ids, peptide_ids);
time_t end_time = time(NULL);
cout << "Search time: " << difftime(end_time, start_time) << " seconds for " << *in << "\n";
}
time_t end_time = time(NULL);
cout << "Total time: " << difftime(end_time, prog_time) << " secconds\n";
return EXECUTION_OK;
}
void CompNovoIdentificationCID::getIdentification(PeptideIdentification & id, const PeakSpectrum & CID_spec)
{
//if (CID_spec.getPrecursors().begin()->getMZ() > 1000.0)
//{
//cerr << "Weight of precursor has been estimated to exceed 2000.0 Da which is the current limit" << endl;
//return;
//}
PeakSpectrum new_CID_spec(CID_spec);
windowMower_(new_CID_spec, 0.3, 1);
Param zhang_param;
zhang_param = zhang_.getParameters();
zhang_param.setValue("tolerance", fragment_mass_tolerance_);
zhang_param.setValue("use_gaussian_factor", "true");
zhang_param.setValue("use_linear_factor", "false");
zhang_.setParameters(zhang_param);
Normalizer normalizer;
Param n_param(normalizer.getParameters());
n_param.setValue("method", "to_one");
normalizer.setParameters(n_param);
normalizer.filterSpectrum(new_CID_spec);
Size charge(2);
double precursor_weight(0); // [M+H]+
if (!CID_spec.getPrecursors().empty())
{
// believe charge of spectrum?
if (CID_spec.getPrecursors().begin()->getCharge() != 0)
{
charge = CID_spec.getPrecursors().begin()->getCharge();
}
else
{
// TODO estimate charge state
}
precursor_weight = CID_spec.getPrecursors().begin()->getMZ() * charge - ((charge - 1) * Constants::PROTON_MASS_U);
}
//cerr << "charge=" << charge << ", [M+H]=" << precursor_weight << endl;
// now delete all peaks that are right of the estimated precursor weight
Size peak_counter(0);
for (PeakSpectrum::ConstIterator it = new_CID_spec.begin(); it != new_CID_spec.end(); ++it, ++peak_counter)
{
if (it->getPosition()[0] > precursor_weight)
{
break;
}
}
if (peak_counter < new_CID_spec.size())
{
new_CID_spec.resize(peak_counter);
}
static double oxonium_mass = EmpiricalFormula("H2O+").getMonoWeight();
Peak1D p;
p.setIntensity(1);
p.setPosition(oxonium_mass);
new_CID_spec.push_back(p);
p.setPosition(precursor_weight);
new_CID_spec.push_back(p);
// add complement to spectrum
/*
for (PeakSpectrum::ConstIterator it1 = CID_spec.begin(); it1 != CID_spec.end(); ++it1)
{
// get m/z of complement
double mz_comp = precursor_weight - it1->getPosition()[0] + Constants::PROTON_MASS_U;
// search if peaks are available that have similar m/z values
Size count(0);
bool found(false);
for (PeakSpectrum::ConstIterator it2 = CID_spec.begin(); it2 != CID_spec.end(); ++it2, ++count)
{
if (fabs(mz_comp - it2->getPosition()[0]) < fragment_mass_tolerance)
{
// add peak intensity to corresponding peak in new_CID_spec
new_CID_spec[count].setIntensity(new_CID_spec[count].getIntensity());
}
}
if (!found)
{
// infer this peak
Peak1D p;
p.setIntensity(it1->getIntensity());
p.setPosition(mz_comp);
new_CID_spec.push_back(p);
}
}*/
CompNovoIonScoringCID ion_scoring;
Param ion_scoring_param(ion_scoring.getParameters());
ion_scoring_param.setValue("fragment_mass_tolerance", fragment_mass_tolerance_);
ion_scoring_param.setValue("precursor_mass_tolerance", precursor_mass_tolerance_);
ion_scoring_param.setValue("decomp_weights_precision", decomp_weights_precision_);
ion_scoring_param.setValue("double_charged_iso_threshold", (double)param_.getValue("double_charged_iso_threshold"));
ion_scoring_param.setValue("max_isotope_to_score", param_.getValue("max_isotope_to_score"));
ion_scoring_param.setValue("max_isotope", max_isotope_);
ion_scoring.setParameters(ion_scoring_param);
Map<double, IonScore> ion_scores;
ion_scoring.scoreSpectrum(ion_scores, new_CID_spec, precursor_weight, charge);
new_CID_spec.sortByPosition();
/*
cerr << "Size of ion_scores " << ion_scores.size() << endl;
for (Map<double, IonScore>::const_iterator it = ion_scores.begin(); it != ion_scores.end(); ++it)
{
cerr << it->first << " " << it->second.score << endl;
}*/
#ifdef WRITE_SCORED_SPEC
PeakSpectrum filtered_spec(new_CID_spec);
filtered_spec.clear();
for (Map<double, CompNovoIonScoringCID::IonScore>::const_iterator it = ion_scores.begin(); it != ion_scores.end(); ++it)
{
Peak1D p;
p.setIntensity(it->second.score);
p.setPosition(it->first);
filtered_spec.push_back(p);
}
DTAFile().store("spec_scored.dta", filtered_spec);
#endif
set<String> sequences;
getDecompositionsDAC_(sequences, 0, new_CID_spec.size() - 1, precursor_weight, new_CID_spec, ion_scores);
#ifdef SPIKE_IN
sequences.insert("AFCVDGEGR");
sequences.insert("APEFAAPWPDFVPR");
sequences.insert("AVKQFEESQGR");
sequences.insert("CCTESLVNR");
sequences.insert("DAFLGSFLYEYSR");
sequences.insert("DAIPENLPPLTADFAEDK");
sequences.insert("DDNKVEDIWSFLSK");
sequences.insert("DDPHACYSTVFDK");
sequences.insert("DEYELLCLDGSR");
sequences.insert("DGAESYKELSVLLPNR");
sequences.insert("DGASCWCVDADGR");
sequences.insert("DLFIPTCLETGEFAR");
sequences.insert("DTHKSEIAHR");
sequences.insert("DVCKNYQEAK");
sequences.insert("EACFAVEGPK");
sequences.insert("ECCHGDLLECADDR");
sequences.insert("EFLGDKFYTVISSLK");
sequences.insert("EFTPVLQADFQK");
sequences.insert("ELFLDSGIFQPMLQGR");
sequences.insert("ETYGDMADCCEK");
sequences.insert("EVGCPSSSVQEMVSCLR");
sequences.insert("EYEATLEECCAK");
sequences.insert("FADLIQSGTFQLHLDSK");
sequences.insert("FFSASCVPGATIEQK");
sequences.insert("FLANVSTVLTSK");
sequences.insert("FLSGSDYAIR");
sequences.insert("FTASCPPSIK");
sequences.insert("GAIEWEGIESGSVEQAVAK");
sequences.insert("GDVAFIQHSTVEENTGGK");
sequences.insert("GEPPSCAEDQSCPSER");
sequences.insert("GEYVPTSLTAR");
sequences.insert("GQEFTITGQKR");
sequences.insert("GTFAALSELHCDK");
sequences.insert("HLVDEPQNLIK");
sequences.insert("HQDCLVTTLQTQPGAVR");
sequences.insert("HTTVNENAPDQK");
sequences.insert("ILDCGSPDTEVR");
sequences.insert("KCPSPCQLQAER");
sequences.insert("KGTEFTVNDLQGK");
sequences.insert("KQTALVELLK");
sequences.insert("KVPQVSTPTLVEVSR");
sequences.insert("LALQFTTNAKR");
sequences.insert("LCVLHEKTPVSEK");
sequences.insert("LFTFHADICTLPDTEK");
sequences.insert("LGEYGFQNALIVR");
sequences.insert("LHVDPENFK");
sequences.insert("LKECCDKPLLEK");
sequences.insert("LKHLVDEPQNLIK");
sequences.insert("LKPDPNTLCDEFK");
sequences.insert("LLGNVLVVVLAR");
sequences.insert("LLVVYPWTQR");
sequences.insert("LRVDPVNFK");
sequences.insert("LTDEELAFPPLSPSR");
sequences.insert("LVNELTEFAK");
sequences.insert("MFLSFPTTK");
sequences.insert("MPCTEDYLSLILNR");
sequences.insert("NAPYSGYSGAFHCLK");
sequences.insert("NECFLSHKDDSPDLPK");
sequences.insert("NEPNKVPACPGSCEEVK");
sequences.insert("NLQMDDFELLCTDGR");
sequences.insert("QAGVQAEPSPK");
sequences.insert("RAPEFAAPWPDFVPR");
sequences.insert("RHPEYAVSVLLR");
sequences.insert("RPCFSALTPDETYVPK");
sequences.insert("RSLLLAPEEGPVSQR");
sequences.insert("SAFPPEPLLCSVQR");
sequences.insert("SAGWNIPIGTLLHR");
sequences.insert("SCWCVDEAGQK");
sequences.insert("SGNPNYPHEFSR");
sequences.insert("SHCIAEVEK");
sequences.insert("SISSGFFECER");
sequences.insert("SKYLASASTMDHAR");
sequences.insert("SLHTLFGDELCK");
sequences.insert("SLLLAPEEGPVSQR");
sequences.insert("SPPQCSPDGAFRPVQCK");
sequences.insert("SREGDPLAVYLK");
sequences.insert("SRQIPQCPTSCER");
sequences.insert("TAGTPVSIPVCDDSSVK");
sequences.insert("TCVADESHAGCEK");
sequences.insert("TQFGCLEGFGR");
sequences.insert("TVMENFVAFVDK");
sequences.insert("TYFPHFDLSHGSAQVK");
sequences.insert("TYMLAFDVNDEK");
sequences.insert("VDEVGGEALGR");
sequences.insert("VDLLIGSSQDDGLINR");
sequences.insert("VEDIWSFLSK");
sequences.insert("VGGHAAEYGAEALER");
sequences.insert("VGTRCCTKPESER");
sequences.insert("VKVDEVGGEALGR");
sequences.insert("VKVDLLIGSSQDDGLINR");
sequences.insert("VLDSFSNGMK");
sequences.insert("VLSAADKGNVK");
sequences.insert("VPQVSTPTLVEVSR");
sequences.insert("VTKCCTESLVNR");
sequences.insert("VVAASDASQDALGCVK");
sequences.insert("VVAGVANALAHR");
sequences.insert("YICDNQDTISSK");
sequences.insert("YLASASTMDHAR");
sequences.insert("YNGVFQECCQAEDK");
#endif
SpectrumAlignmentScore spectra_zhang;
spectra_zhang.setParameters(zhang_param);
vector<PeptideHit> hits;
Size missed_cleavages = param_.getValue("missed_cleavages");
for (set<String>::const_iterator it = sequences.begin(); it != sequences.end(); ++it)
{
Size num_missed = countMissedCleavagesTryptic_(*it);
if (missed_cleavages < num_missed)
{
//cerr << "Two many missed cleavages: " << *it << ", found " << num_missed << ", allowed " << missed_cleavages << endl;
continue;
}
PeakSpectrum CID_sim_spec;
getCIDSpectrum_(CID_sim_spec, *it, charge);
//normalizer.filterSpectrum(CID_sim_spec);
double cid_score = zhang_(CID_sim_spec, CID_spec);
PeptideHit hit;
hit.setScore(cid_score);
hit.setSequence(getModifiedAASequence_(*it));
hit.setCharge((Int)charge); //TODO unify charge interface: int or size?
hits.push_back(hit);
//cerr << getModifiedAASequence_(*it) << " " << cid_score << " " << endl;
}
// rescore the top hits
id.setHits(hits);
id.assignRanks();
hits = id.getHits();
SpectrumAlignmentScore alignment_score;
Param align_param(alignment_score.getParameters());
align_param.setValue("tolerance", fragment_mass_tolerance_);
align_param.setValue("use_linear_factor", "true");
alignment_score.setParameters(align_param);
for (vector<PeptideHit>::iterator it = hits.begin(); it != hits.end(); ++it)
{
//cerr << "Pre: " << it->getRank() << " " << it->getSequence() << " " << it->getScore() << " " << endl;
}
Size number_of_prescoring_hits = param_.getValue("number_of_prescoring_hits");
if (hits.size() > number_of_prescoring_hits)
{
hits.resize(number_of_prescoring_hits);
}
for (vector<PeptideHit>::iterator it = hits.begin(); it != hits.end(); ++it)
{
PeakSpectrum CID_sim_spec;
getCIDSpectrum_(CID_sim_spec, getModifiedStringFromAASequence_(it->getSequence()), charge);
normalizer.filterSpectrum(CID_sim_spec);
//DTAFile().store("sim_specs/" + it->getSequence().toUnmodifiedString() + "_sim_CID.dta", CID_sim_spec);
//double cid_score = spectra_zhang(CID_sim_spec, CID_spec);
double cid_score = alignment_score(CID_sim_spec, CID_spec);
//cerr << "Final: " << it->getSequence() << " " << cid_score << endl;
it->setScore(cid_score);
}
id.setHits(hits);
id.assignRanks();
hits = id.getHits();
for (vector<PeptideHit>::iterator it = hits.begin(); it != hits.end(); ++it)
{
//cerr << "Fin: " << it->getRank() << " " << it->getSequence() << " " << it->getScore() << " " << endl;
}
Size number_of_hits = param_.getValue("number_of_hits");
if (id.getHits().size() > number_of_hits)
{
hits.resize(number_of_hits);
}
id.setHits(hits);
id.assignRanks();
return;
}
void IDDecoyProbability::apply_(vector<PeptideIdentification> & ids, const vector<double> & rev_scores, const vector<double> & fwd_scores, const vector<double> & all_scores)
{
Size number_of_bins(param_.getValue("number_of_bins"));
// normalize distribution to [0, 1]
vector<double> fwd_scores_normalized(number_of_bins, 0.0), rev_scores_normalized(number_of_bins, 0.0), diff_scores(number_of_bins, 0.0), all_scores_normalized(number_of_bins, 0.0);
Transformation_ rev_trafo, fwd_trafo, all_trafo;
normalizeBins_(rev_scores, rev_scores_normalized, rev_trafo);
normalizeBins_(fwd_scores, fwd_scores_normalized, fwd_trafo);
normalizeBins_(all_scores, all_scores_normalized, all_trafo);
// rev scores fitting
vector<DPosition<2> > rev_data;
for (Size i = 0; i < number_of_bins; ++i)
{
DPosition<2> pos;
pos.setX(((double)i) / (double)number_of_bins + 0.0001); // necessary????
pos.setY(rev_scores_normalized[i]);
rev_data.push_back(pos);
#ifdef IDDECOYPROBABILITY_DEBUG
cerr << pos.getX() << " " << pos.getY() << endl;
#endif
}
Math::GammaDistributionFitter gdf;
Math::GammaDistributionFitter::GammaDistributionFitResult result_gamma_1st (1.0, 3.0);
gdf.setInitialParameters(result_gamma_1st);
// TODO heuristic for good start parameters
Math::GammaDistributionFitter::GammaDistributionFitResult result_gamma = gdf.fit(rev_data);
#ifdef IDDECOYPROBABILITY_DEBUG
cerr << gdf.getGnuplotFormula() << endl;
String rev_filename = param_.getValue("rev_filename");
generateDistributionImage_(rev_scores_normalized, gdf.getGnuplotFormula(), rev_filename);
#endif
// generate diffs of distributions
// get the fwd and rev distribution, apply all_trafo and calculate the diff
vector<Size> fwd_bins(number_of_bins, 0), rev_bins(number_of_bins, 0);
double min(all_trafo.min_score), diff(all_trafo.diff_score);
Size max_bin(0);
for (vector<double>::const_iterator it = fwd_scores.begin(); it != fwd_scores.end(); ++it)
{
Size bin = (Size)((*it - min) / diff * (double)(number_of_bins - 1));
++fwd_bins[bin];
if (fwd_bins[bin] > max_bin)
{
max_bin = fwd_bins[bin];
}
}
Size max_reverse_bin(0), max_reverse_bin_value(0);
//min = rev_trafo.min_score;
//diff = rev_trafo.diff_score;
for (vector<double>::const_iterator it = rev_scores.begin(); it != rev_scores.end(); ++it)
{
Size bin = (Size)((*it - min) / diff * (double)number_of_bins);
++rev_bins[bin];
if (rev_bins[bin] > max_bin)
{
max_bin = rev_bins[bin];
}
if (rev_bins[bin] > max_reverse_bin_value)
{
max_reverse_bin = bin;
max_reverse_bin_value = rev_bins[bin];
}
}
#ifdef IDDECOYPROBABILITY_DEBUG
cerr << "Trying to get diff scores" << endl;
#endif
// get diff of fwd and rev
for (Size i = 0; i < number_of_bins; ++i)
{
Size fwd(0), rev(0);
fwd = fwd_bins[i];
rev = rev_bins[i];
if ((double)fwd > (double)(1.3 * rev) && max_reverse_bin < i)
{
diff_scores[i] = (double)(fwd - rev) / (double)max_bin;
}
else
{
diff_scores[i] = 0.0;
}
}
#ifdef IDDECOYPROBABILITY_DEBUG
cerr << "Gauss Fitting values size of diff scores=" << diff_scores.size() << endl;
#endif
// diff scores fitting
vector<DPosition<2> > diff_data;
double gauss_A(0), gauss_x0(0), norm_factor(0);
for (Size i = 0; i < number_of_bins; ++i)
{
DPosition<2> pos;
pos.setX((double)i / (double)number_of_bins);
pos.setY(diff_scores[i]);
if (pos.getY() > gauss_A)
{
gauss_A = pos.getY();
}
gauss_x0 += pos.getX() * pos.getY();
norm_factor += pos.getY();
diff_data.push_back(pos);
}
double gauss_sigma(0);
gauss_x0 /= (double)diff_data.size();
gauss_x0 /= norm_factor;
for (Size i = 0; i <= number_of_bins; ++i)
{
gauss_sigma += fabs(gauss_x0 - (double)i / (double)number_of_bins);
}
gauss_sigma /= (double)diff_data.size();
#ifdef IDDECOYPROBABILITY_DEBUG
cerr << "setting initial parameters: " << endl;
#endif
Math::GaussFitter gf;
Math::GaussFitter::GaussFitResult result_1st(gauss_A, gauss_x0, gauss_sigma);
gf.setInitialParameters(result_1st);
#ifdef IDDECOYPROBABILITY_DEBUG
cerr << "Initial Gauss guess: A=" << gauss_A << ", x0=" << gauss_x0 << ", sigma=" << gauss_sigma << endl;
#endif
//TODO: fail-to-fit correction was done using the GNUPlotFormula. Seemed to be a hack.
//Changed it to try-catch-block but I am not sure if this correction should be made
//at all. Can someone please verify?
Math::GaussFitter::GaussFitResult result_gauss (gauss_A, gauss_x0, gauss_sigma);
try{
result_gauss = gf.fit(diff_data);
}
catch(Exception::UnableToFit& /* e */)
{
result_gauss.A = gauss_A;
result_gauss.x0 = gauss_x0;
result_gauss.sigma = gauss_sigma;
}
// // fit failed?
// if (gf.getGnuplotFormula() == "")
// {
// result_gauss.A = gauss_A;
// result_gauss.x0 = gauss_x0;
// result_gauss.sigma = gauss_sigma;
// }
#ifdef IDDECOYPROBABILITY_DEBUG
cerr << gf.getGnuplotFormula() << endl;
String fwd_filename = param_.getValue("fwd_filename");
if (gf.getGnuplotFormula() == "")
{
String formula("f(x)=" + String(gauss_A) + " * exp(-(x - " + String(gauss_x0) + ") ** 2 / 2 / (" + String(gauss_sigma) + ") ** 2)");
generateDistributionImage_(diff_scores, formula, fwd_filename);
}
else
{
generateDistributionImage_(diff_scores, gf.getGnuplotFormula(), fwd_filename);
}
#endif
#ifdef IDDECOYPROBABILITY_DEBUG
//all_trafo.diff_score + all_trafo.min_score
String gauss_formula("f(x)=" + String(result_gauss.A / all_trafo.max_intensity) + " * exp(-(x - " + String(result_gauss.x0 * all_trafo.diff_score + all_trafo.min_score) + ") ** 2 / 2 / (" + String(result_gauss.sigma * all_trafo.diff_score) + ") ** 2)");
String b_str(result_gamma.b), p_str(result_gamma.p);
String gamma_formula = "g(x)=(" + b_str + " ** " + p_str + ") / gamma(" + p_str + ") * x ** (" + p_str + " - 1) * exp(- " + b_str + " * x)";
generateDistributionImage_(all_scores_normalized, all_trafo, gauss_formula, gamma_formula, (String)param_.getValue("fwd_filename"));
#endif
vector<PeptideIdentification> new_prob_ids;
// calculate the probabilities and write them to the IDs
for (vector<PeptideIdentification>::const_iterator it = ids.begin(); it != ids.end(); ++it)
{
if (it->getHits().size() > 0)
{
vector<PeptideHit> hits;
String score_type = it->getScoreType() + "_score";
for (vector<PeptideHit>::const_iterator pit = it->getHits().begin(); pit != it->getHits().end(); ++pit)
{
PeptideHit hit = *pit;
double score = hit.getScore();
if (!it->isHigherScoreBetter())
{
score = -log10(score);
}
hit.setMetaValue(score_type, hit.getScore());
hit.setScore(getProbability_(result_gamma, rev_trafo, result_gauss, fwd_trafo, score));
hits.push_back(hit);
}
PeptideIdentification id = *it;
id.setHigherScoreBetter(true);
id.setScoreType(id.getScoreType() + "_DecoyProbability");
id.setHits(hits);
new_prob_ids.push_back(id);
}
}
ids = new_prob_ids;
}