unsigned int CAlignmentRefiner::GetBlocksToAlign(unsigned int nBlocks, vector<unsigned int>& blocks, string& msg, bool useExtras) {
bool skip = false;
unsigned int first = 0, last = nBlocks - 1;
blocks.clear();
if (nBlocks == 0) return 0;
CArgs args = GetArgs();
unsigned int nExtra = (useExtras) ? (unsigned int) args.GetNExtra() : 0;
// If specify realignment of all blocks, the default settings are OK.
// Otherwise, use the range specified in the -f and -l flags.
// Recall that the command line takes in a one-based integer; blocks is zero-based.
if (!args["ab"]) {
if (args["f"]) {
first = (unsigned) args["f"].AsInteger() - 1;
}
if (args["l"]) {
last = (unsigned) args["l"].AsInteger() - 1;
}
if (first >= nBlocks) {
first = 0;
}
if (last < first || last >= nBlocks) {
last = nBlocks - 1;
}
}
// If there are any extra arguments provided, they refer to block numbers to freeze.
msg = "\nAligning blocks: ";
for (unsigned int i = first; i <= last; ++i) {
if (nExtra > 0) {
skip = false;
for (size_t extra = 1; extra <= nExtra; ++extra) {
if (args[extra].AsInteger() - 1 == (int) i) {
skip = true;
break;
}
}
if (skip) continue;
}
blocks.push_back(i);
msg.append(NStr::UIntToString(i+1) + " ");
if ((last-first+1)%15 == 0 && first != 0) msg.append("\n");
}
//TERSE_INFO_MESSAGE_CL("message in GetBlocksToAlign:\n" << msg);
return blocks.size();
}
void CAgpValidateApplication::x_ValidateUsingFiles(const CArgs& args, CNcbiOstream* out)
{
if(m_reader.m_is_chr) {
if(m_reader.m_explicit_scaf) {
if(!m_use_xml) {
cout << "===== Reading Chromosome from scaffold AGP =====" << endl;
// second header - for details that are printed below the summary and stats
if(out) *out << "===== Chromosome from scaffold AGP =====" << endl;
}
}
// else: cout << "===== Reading Chromosome from component AGP =====" << endl;
}
else if(m_reader.m_explicit_scaf) {
if(!m_use_xml) {
cout << "===== Reading Scaffold from component AGP =====" << endl;
if(out) *out << "===== Scaffold from component AGP =====" << endl; // header for details that are printed below
}
}
if( 0==(m_ValidationType&VT_Acc) && args["out"].HasValue()) {
CAgpCompSpanSplitter *comp_splitter = new CAgpCompSpanSplitter(&(args["out"].AsOutputFile()));
m_reader.SetRowOutput(comp_splitter);
}
if (args.GetNExtra() == 0) {
x_ValidateFile(cin);
}
else {
SIZE_TYPE num_fasta_files=0;
bool allowFasta = !m_reader.m_explicit_scaf;
for (unsigned int i = 1; i <= args.GetNExtra(); i++) {
m_CurrentFileName = args['#' + NStr::IntToString(i)].AsString();
if(m_CurrentFileName=="-chr") {
if(m_reader.m_is_chr) {
cerr << "Error -- second -chr is not supported.\n";
exit(1);
}
if(!m_reader.m_explicit_scaf) {
cerr << "Error -- -chr after a file, but no preceding -scaf. Expecting:\n"
<< " -scaf Scaffold_AGP_file(s) -chr Chromosome_AGP_file(s)\n";
exit(1);
}
m_reader.PrintTotals(cout, m_use_xml);
m_reader.Reset(true);
pAgpErr->ResetTotals();
if(!m_use_xml) {
cout << "\n===== Reading Chromosome from scaffold AGP =====" << endl;
if(out) *out << "\n===== Chromosome from scaffold AGP =====" << endl;// header for details that are printed below
}
continue;
}
//CNcbiIstream& istr = args['#' + NStr::IntToString(i)].AsInputFile();
CNcbiIfstream istr(m_CurrentFileName.c_str());
if (!istr) {
cerr << "Error -- unable to open file : " << m_CurrentFileName << "\n";
exit (1);
}
char ch=0;
if(allowFasta) {
istr.get(ch); istr.putback(ch);
}
if(ch=='>') {
x_LoadLenFa(istr, m_CurrentFileName);
num_fasta_files++;
}
else {
if(allowFasta && num_fasta_files) x_ReportFastaSeqCount();
if( args.GetNExtra()-num_fasta_files>1 ) pAgpErr->StartFile(m_CurrentFileName);
x_ValidateFile(istr);
allowFasta=false;
}
}
if(num_fasta_files==args.GetNExtra()) {
//cerr << "No AGP files."; exit (1);
if(allowFasta && num_fasta_files) x_ReportFastaSeqCount();
x_ValidateFile(cin);
}
}
}
void CAlignmentRefiner::EchoSettings(ostream& echoStream, bool echoLOO, bool echoBE) {
static string yes = "Yes", no = "No";
CArgs args = GetArgs();
unsigned int nExtra = (unsigned int) args.GetNExtra();
if ((!echoLOO && !echoBE) || (echoLOO && echoBE)) {
echoStream << "Global Refinement Parameters:" << endl;
echoStream << "=================================" << endl;
echoStream << "Number of trials = " << m_nTrials << endl;
echoStream << "Number of cycles per trial = " << m_nCycles << endl;
echoStream << "Alignment score deviation threshold = " << m_scoreDeviationThreshold << endl;
if (nExtra > 0) {
echoStream << "Extra argument(s) freeze " << ((m_loo.extrasAreRows) ? "Row:\n " : "Block:\n ");
for (size_t extra = 1; extra <= nExtra; ++extra) {
echoStream << args[extra].AsInteger() << " ";
}
echoStream << endl;
} else {
echoStream << "No extra arguments that exclude specific rows/blocks from refinement." << endl;
}
// echoStream << "Quiet details mode? " << ((m_quietDetails) ? "ON" : "OFF") << endl;
// echoStream << "Forced threshold (for MC only) = " << m_forcedThreshold << endl;
echoStream << "Quiet mode? " << ((m_quietMode) ? "ON" : "OFF") << endl;
echoStream << endl;
}
if (echoLOO) {
echoStream << "Leave-One_Out parameters:" << endl;
echoStream << "=================================" << endl;
echoStream << "LOO on? " << ((m_loo.doLOO) ? yes : no) << endl;
if (m_loo.doLOO) {
echoStream << "Row selection order: " << RefinerRowSelectorCodeToStr(m_loo.selectorCode) << endl;
echoStream << "Number left out between PSSM recomputation = " << m_loo.lno << endl;
echoStream << "Freeze alignment of rows with structure? " << ((m_loo.fixStructures) ? yes : no) << endl;
echoStream << "Use full sequence or aligned footprint? " << ((m_loo.fullSequence) ? "Full" : "Aligned") << endl;
echoStream << "N-terminal extension allowed = " << m_loo.nExt << endl;
echoStream << "C-terminal extension allowed = " << m_loo.cExt << endl;
echoStream << "Converged after fraction of rows left out do not change score = " << m_loo.sameScoreThreshold << endl;
echoStream << "Random number generator seed = " << m_loo.seed << endl;
echoStream << "LOO loop percentile: longest loop allowed = max initial loop * " << m_loo.percentile << endl;
echoStream << "LOO extension to longest loop allowed = " << m_loo.extension << endl;
echoStream << "LOO absolute maximum longest loop (zero == no max) = " << m_loo.cutoff << endl;
}
echoStream << endl;
}
if (echoBE) {
string algMethod = "Invalid Method";
string columnMethod = algMethod;
switch (m_blockEdit.algMethod) {
case eSimpleExtendAndShrink:
algMethod = "Extend and Shrink";
break;
case eSimpleExtend:
algMethod = "Extend Only";
break;
case eSimpleShrink:
algMethod = "Shrink Only";
break;
case eGreedyExtend:
algMethod = "Greedy Extend Only";
break;
default:
break;
};
switch (m_blockEdit.columnMethod) {
case ePercentAtOrOverThreshold:
columnMethod = "% Rows at or Over Threshold";
break;
case eSumOfScores:
columnMethod = "Sum of Scores";
break;
case eMedianScore:
columnMethod = "Median Score";
break;
case ePercentOfWeightOverThreshold:
columnMethod = "% Score Weight at or Over Threshold";
break;
case eCompoundScorer:
columnMethod = "3.3.3";
if (GetArgs()["be_score"].AsString() != "3.3.3") {
columnMethod = "Compound Scoring";
}
break;
default:
break;
};
echoStream << "Block editing parameters:" << endl;
echoStream << "=================================" << endl;
echoStream << "block editing on? " << ((m_blockEdit.editBlocks) ? yes : no) << endl;
if (m_blockEdit.editBlocks) {
echoStream << "block shrinking on? " << ((m_blockEdit.canShrink) ? yes : no) << endl;
echoStream << "extend first? " << ((m_blockEdit.extendFirst) ? yes : no) << endl;
echoStream << endl;
echoStream << "block editing method = " << algMethod << endl;
echoStream << "column scoring method = " << columnMethod << endl;
echoStream << endl;
// echoStream << "not used: column meth2 = " << m_blockEdit.columnMethod2 << endl << endl;
if (GetArgs()["be_score"].AsString() == "3.3.3") {
echoStream << "(used for 3.3.3 scoring only):" << endl;
echoStream << " median threshold = " << m_blockEdit.median << endl;
echoStream << " negative score fraction = " << m_blockEdit.negScoreFraction << endl;
echoStream << " negative row fraction = " << m_blockEdit.negRowsFraction << endl;
} else {
echoStream << "minimum block size = " << m_blockEdit.minBlockSize << endl;
echoStream << "column-scorer threshold = " << m_blockEdit.columnScorerThreshold << endl;
echoStream << "extension threshold = " << m_blockEdit.extensionThreshold << endl;
echoStream << "shrinkage threshold = " << m_blockEdit.shrinkageThreshold << endl;
}
}
echoStream << endl;
}
}
RefinerResultCode CAlignmentRefiner::ExtractLOOArgs(unsigned int nAlignedBlocks, string& msg) {
int selectionOrder;
unsigned int nBlocksMade, nExtra, extra;
// Get arguments
CArgs args = GetArgs();
RefinerResultCode result = eRefinerResultOK;
msg.erase();
m_loo.doLOO = (!args["no_LOO"]);
m_loo.fixStructures = (args["fix_structs"]);
m_loo.extrasAreRows = (!args["extras_are_blocks"]);
// "selection_order" is mandatory (unless -no_LOO is present) and constrained to {0, 1, 2}.
// number of trials is only relevant for a random selection order.
selectionOrder = (m_loo.doLOO) ? args["selection_order"].AsInteger() : 0;
switch (selectionOrder) {
case 0:
m_loo.selectorCode = eRandomSelectionOrder;
break;
case 1:
m_nTrials = 1;
m_loo.selectorCode = eWorstScoreFirst;
break;
case 2:
m_nTrials = 1;
m_loo.selectorCode = eBestScoreFirst;
break;
};
if (m_loo.doLOO) {
m_loo.fullSequence = (args["fs"]);
m_loo.nExt = (args["nex"]) ? args["nex"].AsInteger() : args["ex"].AsInteger();
m_loo.cExt = (args["cex"]) ? args["cex"].AsInteger() : args["ex"].AsInteger();
m_loo.seed = (args["seed"]) ? args["seed"].AsInteger() : 0;
m_loo.lno = (unsigned int) args["lno"].AsInteger();
m_loo.sameScoreThreshold = args["convSameScore"].AsDouble();
m_loo.percentile = args["p"].AsDouble();
m_loo.extension = (unsigned) args["x"].AsInteger();
m_loo.cutoff = (unsigned) args["c"].AsInteger();
}
if (m_loo.doLOO) {
if (m_loo.extrasAreRows) {
nExtra = (unsigned int) args.GetNExtra();
for (unsigned int i = 1; i <= nExtra; ++i) {
extra = (unsigned int) args[i].AsInteger() - 1;
m_loo.rowsToExclude.push_back(extra);
}
}
// 'false' == don't exclude any blocks using extra cmd line arguments;
nBlocksMade = GetBlocksToAlign(nAlignedBlocks, m_loo.blocks, msg, !m_loo.extrasAreRows);
msg = "Freeze " + NStr::UIntToString(nAlignedBlocks - nBlocksMade) + " blocks in ExtractLOOArgs.\n";
}
return result;
}
int COMSSAMerge::Run()
{
try {
CArgs args = GetArgs();
CRef <COMSSASearch> MySearch(new COMSSASearch);
ESerialDataFormat InFileType(eSerial_Xml), OutFileType(eSerial_Xml);
bool obz2(false); // output bzip2 compressed?
bool ibz2(false); // input bzip2 compressed?
if(args["ox"]) OutFileType = eSerial_Xml;
else if(args["ob"]) OutFileType = eSerial_AsnBinary;
else if(args["ot"]) OutFileType = eSerial_AsnText;
else if(args["obz2"]) {
OutFileType = eSerial_Xml;
obz2 = true;
}
else ERR_POST(Fatal << "output file type not given");
if(args["ix"]) InFileType = eSerial_Xml;
else if(args["ib"]) InFileType = eSerial_AsnBinary;
else if(args["it"]) InFileType = eSerial_AsnText;
else if(args["ibz2"]) {
InFileType = eSerial_Xml;
ibz2 = true;
}
else ERR_POST(Fatal << "input file type not given");
// loop thru input files
if ( args["i"].AsString() != "") {
ifstream is(args["i"].AsString().c_str());
bool Begin(true);
if(!is)
ERR_POST(Fatal << "unable to open input file list " << args["i"].AsString());
while(!is.eof()) {
string iFileName;
NcbiGetline(is, iFileName, "\x0d\x0a");
if(iFileName == "" || is.eof()) continue;
try {
CRef <COMSSASearch> InSearch(new COMSSASearch);
CSearchHelper::ReadCompleteSearch(iFileName, InFileType, ibz2, *InSearch);
// InSearch->ReadCompleteSearch(iFileName, InFileType, ibz2);
if(Begin) {
Begin = false;
MySearch->CopyCMSSearch(InSearch);
}
else {
// add
MySearch->AppendSearch(InSearch);
}
}
catch(CException& e) {
ERR_POST(Fatal << "exception: " << e.what());
return 1;
}
}
}
else if ( args.GetNExtra() ) {
for (size_t extra = 1; extra <= args.GetNExtra(); extra++) {
CRef <COMSSASearch> InSearch(new COMSSASearch);
CSearchHelper::ReadCompleteSearch(args[extra].AsString(), InFileType, ibz2, *InSearch);
//InSearch->ReadCompleteSearch(args[extra].AsString(), InFileType, ibz2);
try {
if(extra == 1) {
// copy
MySearch->CopyCMSSearch(InSearch);
}
else {
// add
MySearch->AppendSearch(InSearch);
}
}
catch(CException& e) {
ERR_POST(Fatal << "exception: " << e.what());
return 1;
}
}
}
// write out the new search
auto_ptr <CNcbiOfstream> raw_out;
auto_ptr <CCompressionOStream> compress_out;
auto_ptr <CObjectOStream> txt_out;
if( obz2 ) {
raw_out.reset(new CNcbiOfstream(args["o"].AsString().c_str()));
compress_out.reset( new CCompressionOStream (*raw_out,
new CBZip2StreamCompressor(),
CCompressionStream::fOwnProcessor));
txt_out.reset(CObjectOStream::Open(OutFileType, *compress_out));
}
else {
txt_out.reset(CObjectOStream::Open(args["o"].AsString().c_str(), OutFileType));
}
// auto_ptr <CObjectOStream> txt_out(
// CObjectOStream::Open(args["o"].AsString(), OutFileType));
if(txt_out.get()) {
SetUpOutputFile(txt_out.get(), OutFileType);
if (args["sw"]) {
txt_out->Write(ObjectInfo(*(*MySearch->SetResponse().begin())));
}
else {
txt_out->Write(ObjectInfo(*MySearch));
}
txt_out->Flush();
txt_out->Close();
}
} catch (NCBI_NS_STD::exception& e) {
ERR_POST(Fatal << "Exception in COMSSAMerge::Run: " << e.what());
}
return 0;
}