XHMM::xhmmInputManager::LoadedReadDepths
XHMM::xhmmInputManager::loadReadDepthsFromFile(string readDepthFile, IntervalSet* excludeTargets, StringSet* excludeTargetChromosomes, StringSet* excludeSamples,
const ullint minTargetSize, const ullint maxTargetSize) {
HMM_PP::DoubleMat* rdMat = MatrixReader<double>::readMatrixFromFile(readDepthFile);
StringSet* excludedSamples = new StringSet();
IntervalSet* excludedTargets = new IntervalSet();
set<ullint>* excludeSampleIndices = new set<ullint>();
if (excludeSamples != NULL) {
for (ullint row = 0; row < rdMat->nrow(); ++row) {
string samp = rdMat->rowName(row);
if (excludeSamples->find(samp) != excludeSamples->end()) {
cerr << "Excluded sample " << samp << endl;
excludedSamples->insert(samp);
excludeSampleIndices->insert(row);
}
}
}
set<ullint>* excludeTargetIndices = new set<ullint>();
if (excludeTargets != NULL || excludeTargetChromosomes != NULL || minTargetSize > 0 || maxTargetSize < ULLINT_INFINITY) {
for (ullint j = 0; j < rdMat->ncol(); ++j) {
const Interval targJ(rdMat->colName(j));
const ullint targLen = targJ.span();
bool targLenFails = (targLen < minTargetSize || targLen > maxTargetSize);
if ((excludeTargets != NULL && excludeTargets->find(targJ) != excludeTargets->end()) ||
(excludeTargetChromosomes != NULL && excludeTargetChromosomes->find(targJ.getChr()) != excludeTargetChromosomes->end()) ||
targLenFails) {
cerr << "Excluded target " << targJ;
if (targLenFails)
cerr << " of length " << targLen;
cerr << endl;
excludeTargetIndices->insert(j);
excludedTargets->insert(targJ);
}
}
}
if (!excludeSampleIndices->empty() || !excludeTargetIndices->empty()) {
HMM_PP::DoubleMat* newRdMat = rdMat->deleteRowsAndColumns(excludeSampleIndices, excludeTargetIndices);
delete rdMat;
rdMat = newRdMat;
}
delete excludeSampleIndices;
delete excludeTargetIndices;
return LoadedReadDepths(rdMat, excludedTargets, excludedSamples);
}
void XHMM::ReadDepthMatrixLoader::readTargets() {
stringstream* lineStream = new stringstream(*_header);
delete _header;
_header = NULL;
Interval* prevTarget = NULL;
while (*lineStream && !lineStream->eof()) {
string targetString;
*lineStream >> targetString;
if (!*lineStream)
throw new Exception("Data input stream failed while reading target " + targetString);
Interval* curTarget = new Interval(targetString);
if (_chrStopInds->find(curTarget->getChr()) != _chrStopInds->end())
throw new Exception("MUST provide targets in order and GROUPED BY CHROMOSOME: target " + curTarget->intervalString() + " reverts to chromosome " + curTarget->getChr() + ", which was interrupted by targets in other chromosomes");
_targets->push_back(*curTarget);
if (prevTarget == NULL || curTarget->getChr() != prevTarget->getChr()) // curTarget is a new chromosome
(*_chrStartInds)[curTarget->getChr()] = _targets->size() - 1;
if (prevTarget != NULL) {
if (curTarget->getChr() != prevTarget->getChr()) // switched to a new chromosome, so can never go back to prevTarget->getChr()
(*_chrStopInds)[prevTarget->getChr()] = _targets->size() - 2; // subtract 2 (and not 1) since already added curTarget
else if (!(curTarget->getBp1() > prevTarget->getBp2()))
throw new Exception("MUST provide NON-OVERLAPPING targets IN ORDER: target " + curTarget->intervalString() + " either overlaps with or precedes previous target " + prevTarget->intervalString());
delete prevTarget;
}
prevTarget = curTarget;
}
if (prevTarget != NULL) {
(*_chrStopInds)[prevTarget->getChr()] = _targets->size() - 1; // since now prevTarget points to the LAST target, which ends its chromosome
delete prevTarget;
}
for (map<string, uint>::const_iterator stopIt = _chrStopInds->begin(); stopIt != _chrStopInds->end(); ++stopIt)
_listAllChrStopInds->insert(stopIt->second);
delete lineStream;
}