void ConsensusMapNormalizerAlgorithmQuantile::extractIntensityVectors(const ConsensusMap& map, vector<vector<double> >& out_intensities)
{
//reserve space for out_intensities (unequal vector lengths, 0-features omitted)
Size number_of_maps = map.getColumnHeaders().size();
out_intensities.clear();
out_intensities.resize(number_of_maps);
for (UInt i = 0; i < number_of_maps; i++)
{
ConsensusMap::ColumnHeaders::const_iterator it = map.getColumnHeaders().find(i);
if (it == map.getColumnHeaders().end()) throw Exception::ElementNotFound(__FILE__, __LINE__, OPENMS_PRETTY_FUNCTION, String(i));
out_intensities[i].reserve(it->second.size);
}
//fill out_intensities
ConsensusMap::ConstIterator cf_it;
for (cf_it = map.begin(); cf_it != map.end(); ++cf_it)
{
ConsensusFeature::HandleSetType::const_iterator f_it;
for (f_it = cf_it->getFeatures().begin(); f_it != cf_it->getFeatures().end(); ++f_it)
{
out_intensities[f_it->getMapIndex()].push_back(f_it->getIntensity());
}
}
}
void ConsensusMapNormalizerAlgorithmQuantile::setNormalizedIntensityValues(const vector<vector<double> >& feature_ints, ConsensusMap& map)
{
//assumes the input map and feature_ints are in the same order as in the beginning,
//although feature_ints has normalized values now (but the same ranks as before)
Size number_of_maps = map.getColumnHeaders().size();
ConsensusMap::ConstIterator cf_it;
vector<Size> progress_indices(number_of_maps);
for (cf_it = map.begin(); cf_it != map.end(); ++cf_it)
{
ConsensusFeature::HandleSetType::const_iterator f_it;
for (f_it = cf_it->getFeatures().begin(); f_it != cf_it->getFeatures().end(); ++f_it)
{
Size map_idx = f_it->getMapIndex();
double intensity = feature_ints[map_idx][progress_indices[map_idx]++];
f_it->asMutable().setIntensity(intensity);
}
}
}
Param p;
p.setValue("rt_estimate","false");
p.setValue("rt_pair_dist",0.4);
p.setValue("rt_dev_low",1.0);
p.setValue("rt_dev_high",2.0);
p.setValue("mz_pair_dists",ListUtils::create<double>(4.0));
p.setValue("mz_dev",0.6);
pm.setParameters(p);
ConsensusMap output;
TEST_EXCEPTION(Exception::IllegalArgument,pm.run(vector<ConsensusMap>(),output));
vector<ConsensusMap> input(1);
MapConversion::convert(5,features,input[0]);
output.getColumnHeaders()[5].label = "light";
output.getColumnHeaders()[5].filename = "filename";
output.getColumnHeaders()[8] = output.getColumnHeaders()[5];
output.getColumnHeaders()[8].label = "heavy";
pm.run(input,output);
TEST_EQUAL(output.size(),1);
ABORT_IF(output.size()!=1)
TEST_REAL_SIMILAR(output[0].begin()->getMZ(),1.0f);
TEST_REAL_SIMILAR(output[0].begin()->getRT(),1.0f);
TEST_REAL_SIMILAR(output[0].rbegin()->getMZ(),5.0f);
TEST_REAL_SIMILAR(output[0].rbegin()->getRT(),1.5f);
TEST_REAL_SIMILAR(output[0].getQuality(),0.959346);
TEST_EQUAL(output[0].getCharge(),1);
//test automated RT parameter estimation
LabeledPairFinder pm2;
ExitCodes main_(int, const char **) override
{
FeatureGroupingAlgorithmUnlabeled * algorithm = new FeatureGroupingAlgorithmUnlabeled();
//-------------------------------------------------------------
// parameter handling
//-------------------------------------------------------------
StringList ins;
ins = getStringList_("in");
String out = getStringOption_("out");
//-------------------------------------------------------------
// check for valid input
//-------------------------------------------------------------
// check if all input files have the correct type
FileTypes::Type file_type = FileHandler::getType(ins[0]);
for (Size i = 0; i < ins.size(); ++i)
{
if (FileHandler::getType(ins[i]) != file_type)
{
writeLog_("Error: All input files must be of the same type!");
return ILLEGAL_PARAMETERS;
}
}
//-------------------------------------------------------------
// set up algorithm
//-------------------------------------------------------------
Param algorithm_param = getParam_().copy("algorithm:", true);
writeDebug_("Used algorithm parameters", algorithm_param, 3);
algorithm->setParameters(algorithm_param);
Size reference_index(0);
//-------------------------------------------------------------
// perform grouping
//-------------------------------------------------------------
// load input
ConsensusMap out_map;
StringList ms_run_locations;
if (file_type == FileTypes::FEATUREXML)
{
// use map with highest number of features as reference:
Size max_count(0);
FeatureXMLFile f;
for (Size i = 0; i < ins.size(); ++i)
{
Size s = f.loadSize(ins[i]);
if (s > max_count)
{
max_count = s;
reference_index = i;
}
}
// Load reference map and input it to the algorithm
UInt64 ref_id;
Size ref_size;
std::vector<PeptideIdentification> ref_pepids;
std::vector<ProteinIdentification> ref_protids;
{
FeatureMap map_ref;
FeatureXMLFile f_fxml_tmp;
f_fxml_tmp.getOptions().setLoadConvexHull(false);
f_fxml_tmp.getOptions().setLoadSubordinates(false);
f_fxml_tmp.load(ins[reference_index], map_ref);
algorithm->setReference(reference_index, map_ref);
ref_id = map_ref.getUniqueId();
ref_size = map_ref.size();
ref_pepids = map_ref.getUnassignedPeptideIdentifications();
ref_protids = map_ref.getProteinIdentifications();
}
ConsensusMap dummy;
// go through all input files and add them to the result one by one
for (Size i = 0; i < ins.size(); ++i)
{
FeatureXMLFile f_fxml_tmp;
FeatureMap tmp_map;
f_fxml_tmp.getOptions().setLoadConvexHull(false);
f_fxml_tmp.getOptions().setLoadSubordinates(false);
f_fxml_tmp.load(ins[i], tmp_map);
// copy over information on the primary MS run
StringList ms_runs;
tmp_map.getPrimaryMSRunPath(ms_runs);
ms_run_locations.insert(ms_run_locations.end(), ms_runs.begin(), ms_runs.end());
if (i != reference_index)
{
algorithm->addToGroup(i, tmp_map);
// store some meta-data about the maps in the "dummy" object -> try to
// keep the same order as they were given in the input independent of
// which map is the reference.
dummy.getColumnHeaders()[i].filename = ins[i];
dummy.getColumnHeaders()[i].size = tmp_map.size();
dummy.getColumnHeaders()[i].unique_id = tmp_map.getUniqueId();
// add protein identifications to result map
dummy.getProteinIdentifications().insert(
dummy.getProteinIdentifications().end(),
tmp_map.getProteinIdentifications().begin(),
tmp_map.getProteinIdentifications().end());
// add unassigned peptide identifications to result map
dummy.getUnassignedPeptideIdentifications().insert(
dummy.getUnassignedPeptideIdentifications().end(),
tmp_map.getUnassignedPeptideIdentifications().begin(),
tmp_map.getUnassignedPeptideIdentifications().end());
}
else
{
// copy the meta-data from the refernce map
dummy.getColumnHeaders()[i].filename = ins[i];
dummy.getColumnHeaders()[i].size = ref_size;
dummy.getColumnHeaders()[i].unique_id = ref_id;
// add protein identifications to result map
dummy.getProteinIdentifications().insert(
dummy.getProteinIdentifications().end(),
ref_protids.begin(),
ref_protids.end());
// add unassigned peptide identifications to result map
dummy.getUnassignedPeptideIdentifications().insert(
dummy.getUnassignedPeptideIdentifications().end(),
ref_pepids.begin(),
ref_pepids.end());
}
}
// get the resulting map
out_map = algorithm->getResultMap();
//
// Copy back meta-data (Protein / Peptide ids / File descriptions)
//
// add protein identifications to result map
out_map.getProteinIdentifications().insert(
out_map.getProteinIdentifications().end(),
dummy.getProteinIdentifications().begin(),
dummy.getProteinIdentifications().end());
// add unassigned peptide identifications to result map
out_map.getUnassignedPeptideIdentifications().insert(
out_map.getUnassignedPeptideIdentifications().end(),
dummy.getUnassignedPeptideIdentifications().begin(),
dummy.getUnassignedPeptideIdentifications().end());
out_map.setColumnHeaders(dummy.getColumnHeaders());
// canonical ordering for checking the results, and the ids have no real meaning anyway
// the way this was done in DelaunayPairFinder and StablePairFinder
// -> the same ordering as FeatureGroupingAlgorithmUnlabeled::group applies!
out_map.sortByMZ();
out_map.updateRanges();
}
else
{
vector<ConsensusMap> maps(ins.size());
ConsensusXMLFile f;
for (Size i = 0; i < ins.size(); ++i)
{
f.load(ins[i], maps[i]);
StringList ms_runs;
maps[i].getPrimaryMSRunPath(ms_runs);
ms_run_locations.insert(ms_run_locations.end(), ms_runs.begin(), ms_runs.end());
}
// group
algorithm->FeatureGroupingAlgorithm::group(maps, out_map);
// set file descriptions:
bool keep_subelements = getFlag_("keep_subelements");
if (!keep_subelements)
{
for (Size i = 0; i < ins.size(); ++i)
{
out_map.getColumnHeaders()[i].filename = ins[i];
out_map.getColumnHeaders()[i].size = maps[i].size();
out_map.getColumnHeaders()[i].unique_id = maps[i].getUniqueId();
}
}
else
{
// components of the output map are not the input maps themselves, but
// the components of the input maps:
algorithm->transferSubelements(maps, out_map);
}
}
// assign unique ids
out_map.applyMemberFunction(&UniqueIdInterface::setUniqueId);
// annotate output with data processing info
addDataProcessing_(out_map, getProcessingInfo_(DataProcessing::FEATURE_GROUPING));
out_map.setPrimaryMSRunPath(ms_run_locations);
// write output
ConsensusXMLFile().store(out, out_map);
// some statistics
map<Size, UInt> num_consfeat_of_size;
for (ConsensusMap::const_iterator cmit = out_map.begin(); cmit != out_map.end(); ++cmit)
{
++num_consfeat_of_size[cmit->size()];
}
LOG_INFO << "Number of consensus features:" << endl;
for (map<Size, UInt>::reverse_iterator i = num_consfeat_of_size.rbegin(); i != num_consfeat_of_size.rend(); ++i)
{
LOG_INFO << " of size " << setw(2) << i->first << ": " << setw(6) << i->second << endl;
}
LOG_INFO << " total: " << setw(6) << out_map.size() << endl;
delete algorithm;
return EXECUTION_OK;
}
void group(const vector< FeatureMap >&, ConsensusMap& map) override
{
map.getColumnHeaders()[0].filename = "bla";
map.getColumnHeaders()[0].size = 5;
}
FGA* nullPointer = nullptr;
START_SECTION((FeatureGroupingAlgorithm()))
ptr = new FGA();
TEST_NOT_EQUAL(ptr, nullPointer)
END_SECTION
START_SECTION((virtual ~FeatureGroupingAlgorithm()))
delete ptr;
END_SECTION
START_SECTION((virtual void group(const vector< FeatureMap > &maps, ConsensusMap &out)=0))
FGA fga;
vector< FeatureMap > in;
ConsensusMap map;
fga.group(in,map);
TEST_EQUAL(map.getColumnHeaders()[0].filename, "bla")
END_SECTION
START_SECTION((static void registerChildren()))
{
TEST_STRING_EQUAL(Factory<FeatureGroupingAlgorithm>::registeredProducts()[0],FeatureGroupingAlgorithmLabeled::getProductName());
TEST_STRING_EQUAL(Factory<FeatureGroupingAlgorithm>::registeredProducts()[1],FeatureGroupingAlgorithmUnlabeled::getProductName());
TEST_EQUAL(Factory<FeatureGroupingAlgorithm>::registeredProducts().size(), 4)
}
END_SECTION
START_SECTION((void transferSubelements(const vector<ConsensusMap>& maps, ConsensusMap& out) const))
{
vector<ConsensusMap> maps(2);
maps[0].getColumnHeaders()[0].filename = "file1";
maps[0].getColumnHeaders()[0].size = 1;
Param p;
p.setValue("rt_estimate","false");
p.setValue("rt_pair_dist",0.4);
p.setValue("rt_dev_low",1.0);
p.setValue("rt_dev_high",2.0);
p.setValue("mz_pair_dists",ListUtils::create<double>(4.0));
p.setValue("mz_dev",0.6);
fga.setParameters(p);
//test exception (no file name set in out)
TEST_EXCEPTION(Exception::IllegalArgument, fga.group(in,out));
out.getColumnHeaders()[5].label = "light";
out.getColumnHeaders()[5].filename = "filename";
out.getColumnHeaders()[8] = out.getColumnHeaders()[5];
out.getColumnHeaders()[8].label = "heavy";
fga.group(in,out);
TEST_EQUAL(out.size(),1)
TEST_REAL_SIMILAR(out[0].getQuality(),0.959346);
TEST_EQUAL(out[0].size(),2)
ConsensusFeature::HandleSetType::const_iterator it = out[0].begin();
TEST_REAL_SIMILAR(it->getMZ(),1.0f);
TEST_REAL_SIMILAR(it->getRT(),1.0f);
++it;
TEST_REAL_SIMILAR(it->getMZ(),5.0f);
TEST_REAL_SIMILAR(it->getRT(),1.5f);
END_SECTION
