MasterDependentAlignment::MasterDependentAlignment(StructureBase *parent, const Sequence *masterSequence,
const ncbi::objects::CSeq_align& seqAlign) :
StructureBase(parent), master(masterSequence), dependent(NULL)
{
// resize alignment and block vector
masterToDependent.resize(master->Length(), -1);
blockStructure.resize(master->Length(), -1);
// find dependent sequence for this alignment, and order (master or dependent first)
const CSeq_id& frontSeqId = seqAlign.GetSegs().IsDendiag() ?
seqAlign.GetSegs().GetDendiag().front()->GetIds().front().GetObject() :
seqAlign.GetSegs().GetDenseg().GetIds().front().GetObject();
const CSeq_id& backSeqId = seqAlign.GetSegs().IsDendiag() ?
seqAlign.GetSegs().GetDendiag().front()->GetIds().back().GetObject() :
seqAlign.GetSegs().GetDenseg().GetIds().back().GetObject();
bool masterFirst = true;
SequenceSet::SequenceList::const_iterator
s, se = master->parentSet->sequenceSet->sequences.end();
for (s=master->parentSet->sequenceSet->sequences.begin(); s!=se; ++s) {
if (master->identifier->MatchesSeqId(frontSeqId) &&
(*s)->identifier->MatchesSeqId(backSeqId)) {
break;
} else if ((*s)->identifier->MatchesSeqId(frontSeqId) &&
master->identifier->MatchesSeqId(backSeqId)) {
masterFirst = false;
break;
}
}
if (s == se) {
ERRORMSG("MasterDependentAlignment::MasterDependentAlignment() - couldn't find matching sequences; "
<< "both " << frontSeqId.AsFastaString() << " and "
<< backSeqId.AsFastaString() << " must be in the sequence list for this file!");
return;
} else {
dependent = *s;
}
unsigned int i, blockNum = 0;
int masterRes, dependentRes;
// unpack dendiag alignment
if (seqAlign.GetSegs().IsDendiag()) {
CSeq_align::C_Segs::TDendiag::const_iterator d , de = seqAlign.GetSegs().GetDendiag().end();
for (d=seqAlign.GetSegs().GetDendiag().begin(); d!=de; ++d, ++blockNum) {
const CDense_diag& block = d->GetObject();
if (block.GetDim() != 2 || block.GetIds().size() != 2 || block.GetStarts().size() != 2) {
ERRORMSG("MasterDependentAlignment::MasterDependentAlignment() - \n"
"incorrect dendiag block dimensions");
return;
}
// make sure identities of master and dependent sequences match in each block
if ((masterFirst &&
(!master->identifier->MatchesSeqId(block.GetIds().front().GetObject()) ||
!dependent->identifier->MatchesSeqId(block.GetIds().back().GetObject()))) ||
(!masterFirst &&
(!master->identifier->MatchesSeqId(block.GetIds().back().GetObject()) ||
!dependent->identifier->MatchesSeqId(block.GetIds().front().GetObject())))) {
ERRORMSG("MasterDependentAlignment::MasterDependentAlignment() - "
"mismatched Seq-id in dendiag block");
return;
}
// finally, actually unpack the data into the alignment vector
for (i=0; i<block.GetLen(); ++i) {
if (masterFirst) {
masterRes = block.GetStarts().front() + i;
dependentRes = block.GetStarts().back() + i;
} else {
masterRes = block.GetStarts().back() + i;
dependentRes = block.GetStarts().front() + i;
}
if (masterRes < 0 || masterRes >= (int)master->Length() || dependentRes < 0 || dependentRes >= (int)dependent->Length()) {
ERRORMSG("MasterDependentAlignment::MasterDependentAlignment() - seqloc in dendiag block > length of sequence!");
return;
}
masterToDependent[masterRes] = dependentRes;
blockStructure[masterRes] = blockNum;
}
}
}
// unpack denseg alignment
else if (seqAlign.GetSegs().IsDenseg()) {
const CDense_seg& block = seqAlign.GetSegs().GetDenseg();
if (!block.IsSetDim() || block.GetDim() != 2 ||
block.GetIds().size() != 2 ||
(int)block.GetStarts().size() != 2 * block.GetNumseg() ||
(int)block.GetLens().size() != block.GetNumseg()) {
ERRORMSG("MasterDependentAlignment::MasterDependentAlignment() - \n"
"incorrect denseg block dimension");
return;
}
// make sure identities of master and dependent sequences match in each block
if ((masterFirst &&
(!master->identifier->MatchesSeqId(block.GetIds().front().GetObject()) ||
!dependent->identifier->MatchesSeqId(block.GetIds().back().GetObject()))) ||
(!masterFirst &&
(!master->identifier->MatchesSeqId(block.GetIds().back().GetObject()) ||
!dependent->identifier->MatchesSeqId(block.GetIds().front().GetObject())))) {
ERRORMSG("MasterDependentAlignment::MasterDependentAlignment() - \n"
"mismatched Seq-id in denseg block");
return;
}
// finally, actually unpack the data into the alignment vector
CDense_seg::TStarts::const_iterator starts = block.GetStarts().begin();
CDense_seg::TLens::const_iterator lens, le = block.GetLens().end();
for (lens=block.GetLens().begin(); lens!=le; ++lens) {
if (masterFirst) {
masterRes = *(starts++);
dependentRes = *(starts++);
} else {
dependentRes = *(starts++);
masterRes = *(starts++);
}
if (masterRes != -1 && dependentRes != -1) { // skip gaps
if ((masterRes + *lens - 1) >= master->Length() ||
(dependentRes + *lens - 1) >= dependent->Length()) {
ERRORMSG("MasterDependentAlignment::MasterDependentAlignment() - \n"
"seqloc in denseg block > length of sequence!");
return;
}
for (i=0; i<*lens; ++i) {
masterToDependent[masterRes + i] = dependentRes + i;
blockStructure[masterRes + i] = blockNum;
}
++blockNum; // a "block" of a denseg is an aligned (non-gap) segment
}
}
}
//TESTMSG("got alignment for dependent gi " << dependent->identifier->gi);
}
MasterSlaveAlignment::MasterSlaveAlignment(
const SequenceSet *sequenceSet,
const Sequence *masterSequence,
const CSeq_align& seqAlign) :
status(CAV_ERROR_PAIRWISE), master(masterSequence), slave(NULL)
{
// resize alignment and block vector
masterToSlave.resize(master->Length(), -1);
SequenceSet::SequenceList::const_iterator
s = sequenceSet->sequences.begin(),
se = sequenceSet->sequences.end();
// find slave sequence for this alignment, and order (master or slave first)
const SeqIdList& sids = seqAlign.GetSegs().IsDendiag() ?
seqAlign.GetSegs().GetDendiag().front()->GetIds() :
seqAlign.GetSegs().GetDenseg().GetIds();
bool masterFirst = true;
for (; s!=se; ++s) {
if ((*s)->Matches(master->bioseqASN->GetId())) continue;
if ((*s)->Matches(sids.back().GetObject())) {
break;
} else if ((*s)->Matches(sids.front().GetObject())) {
masterFirst = false;
break;
}
}
if (s == se) {
// special case of master seq. aligned to itself
if (master->Matches(sids.back().GetObject()) && master->Matches(sids.front().GetObject())) {
slave = master;
} else {
ERR_POST_X(9, Error << "MasterSlaveAlignment::MasterSlaveAlignment() - \n"
"couldn't find matching unaligned slave sequence");
return;
}
} else {
slave = *s;
}
unsigned int i;
int masterRes, slaveRes;
// unpack dendiag alignment
if (seqAlign.GetSegs().IsDendiag()) {
CSeq_align::C_Segs::TDendiag::const_iterator d , de = seqAlign.GetSegs().GetDendiag().end();
for (d=seqAlign.GetSegs().GetDendiag().begin(); d!=de; ++d) {
const CDense_diag& block = d->GetObject();
if (!block.IsSetDim() || block.GetDim() != 2 ||
block.GetIds().size() != 2 ||
block.GetStarts().size() != 2) {
ERR_POST_X(10, Error << "MasterSlaveAlignment::MasterSlaveAlignment() - \n"
"incorrect dendiag block dimensions");
return;
}
// make sure identities of master and slave sequences match in each block
if ((masterFirst &&
(!master->Matches(block.GetIds().front().GetObject()) ||
!slave->Matches(block.GetIds().back().GetObject()))) ||
(!masterFirst &&
(!master->Matches(block.GetIds().back().GetObject()) ||
!slave->Matches(block.GetIds().front().GetObject())))) {
ERR_POST_X(11, Error << "MasterSlaveAlignment::MasterSlaveAlignment() - \n"
"mismatched Seq-id in dendiag block");
return;
}
// finally, actually unpack the data into the alignment vector
for (i=0; i<block.GetLen(); ++i) {
if (masterFirst) {
masterRes = block.GetStarts().front() + i;
slaveRes = block.GetStarts().back() + i;
} else {
masterRes = block.GetStarts().back() + i;
slaveRes = block.GetStarts().front() + i;
}
if (masterRes >= (int) master->Length() || slaveRes >= (int) slave->Length()) {
ERR_POST_X(12, Critical << "MasterSlaveAlignment::MasterSlaveAlignment() - \n"
"seqloc in dendiag block > length of sequence!");
return;
}
masterToSlave[masterRes] = slaveRes;
}
}
}
// unpack denseg alignment
else if (seqAlign.GetSegs().IsDenseg()) {
const CDense_seg& block = seqAlign.GetSegs().GetDenseg();
if (!block.IsSetDim() || block.GetDim() != 2 ||
block.GetIds().size() != 2 ||
block.GetStarts().size() != ((unsigned int ) 2 * block.GetNumseg()) ||
block.GetLens().size() != ((unsigned int ) block.GetNumseg()))
{
ERR_POST_X(13, Error << "MasterSlaveAlignment::MasterSlaveAlignment() - \n"
"incorrect denseg block dimension");
return;
}
// make sure identities of master and slave sequences match in each block
if ((masterFirst &&
(!master->Matches(block.GetIds().front().GetObject()) ||
!slave->Matches(block.GetIds().back().GetObject()))) ||
(!masterFirst &&
(!master->Matches(block.GetIds().back().GetObject()) ||
!slave->Matches(block.GetIds().front().GetObject())))) {
ERR_POST_X(14, Error << "MasterSlaveAlignment::MasterSlaveAlignment() - \n"
"mismatched Seq-id in denseg block");
return;
}
// finally, actually unpack the data into the alignment vector
CDense_seg::TStarts::const_iterator starts = block.GetStarts().begin();
CDense_seg::TLens::const_iterator lens, le = block.GetLens().end();
for (lens=block.GetLens().begin(); lens!=le; ++lens) {
if (masterFirst) {
masterRes = *(starts++);
slaveRes = *(starts++);
} else {
slaveRes = *(starts++);
masterRes = *(starts++);
}
if (masterRes != -1 && slaveRes != -1) { // skip gaps
if ((masterRes + *lens - 1) >= master->Length() ||
(slaveRes + *lens - 1) >= slave->Length()) {
ERR_POST_X(15, Critical << "MasterSlaveAlignment::MasterSlaveAlignment() - \n"
"seqloc in denseg block > length of sequence!");
return;
}
for (i=0; i<*lens; ++i)
masterToSlave[masterRes + i] = slaveRes + i;
}
}
}
status = CAV_SUCCESS;
}