// ----------------------------------------------------------------------------
ESpecType SpecType(
const string& spectype )
// ----------------------------------------------------------------------------
{
static map<string, ESpecType> typemap;
if ( typemap.empty() ) {
typemap["Integer"] = eType_Integer;
typemap["Float"] = eType_Float;
typemap["Flag"] = eType_Flag;
typemap["Character"] = eType_Character;
typemap["String"] = eType_String;
}
try {
return typemap[spectype];
}
catch( ... ) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Warning,
0,
"CVcfReader::xProcessMetaLineInfo: Unrecognized line or record type.",
ILineError::eProblem_GeneralParsingError) );
pErr->Throw();
return eType_String;
}
};
// ----------------------------------------------------------------------------
bool
CVcfReader::xProcessMetaLineFilter(
const string& line,
CRef<CSeq_annot> pAnnot,
IMessageListener* pEC)
// ----------------------------------------------------------------------------
{
const string prefix = "##FILTER=<";
const string postfix = ">";
if ( ! NStr::StartsWith( line, prefix ) || ! NStr::EndsWith( line, postfix ) ) {
return false;
}
try {
vector<string> fields;
string key, id, description;
string info = line.substr(
prefix.length(), line.length() - prefix.length() - postfix.length() );
NStr::Tokenize( info, ",", fields );
NStr::SplitInTwo( fields[0], "=", key, id );
if ( key != "ID" ) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
0,
"CVcfReader::xProcessMetaLineInfo: ##FILTER with bad or missing \"ID\".",
ILineError::eProblem_BadFilterLine) );
pErr->Throw();
}
NStr::SplitInTwo( fields[1], "=", key, description );
if ( key != "Description" ) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
0,
"CVcfReader::xProcessMetaLineInfo: ##FILTER with bad or missing \"Description\".",
ILineError::eProblem_BadFilterLine) );
pErr->Throw();
}
m_FilterSpecs[id] = CVcfFilterSpec( id, description );
}
catch (CObjReaderLineException& err) {
ProcessError(err, pEC);
}
return true;
}
// ---------------------------------------------------------------------------
void
CGtfReader::ReadSeqAnnots(
TAnnots& annots,
ILineReader& lr,
ILineErrorListener* pEC)
// ----------------------------------------------------------------------------
{
xProgressInit(lr);
string line;
while (xGetLine(lr, line)) {
if (IsCanceled()) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Info,
0,
"Reader stopped by user.",
ILineError::eProblem_ProgressInfo));
ProcessError(*pErr, pEC);
annots.clear();
return;
}
xReportProgress(pEC);
try {
if (xIsTrackTerminator(line)) {
continue;
}
if (x_ParseBrowserLineGff(line, m_CurrentBrowserInfo)) {
continue;
}
if (xParseTrackLine(line, pEC)) {
continue;
}
if (x_ParseFeatureGff(line, annots, pEC)) {
continue;
}
}
catch(CObjReaderLineException& err) {
err.SetLineNumber(m_uLineNumber);
}
}
}
// ----------------------------------------------------------------------------
CRef< CSeq_annot >
CVcfReader::ReadSeqAnnot(
ILineReader& lr,
IMessageListener* pEC )
// ----------------------------------------------------------------------------
{
CRef< CSeq_annot > annot( new CSeq_annot );
CRef< CAnnot_descr > desc( new CAnnot_descr );
annot->SetDesc( *desc );
annot->SetData().SetFtable();
m_Meta.Reset( new CAnnotdesc );
m_Meta->SetUser().SetType().SetStr( "vcf-meta-info" );
while ( ! lr.AtEOF() ) {
m_uLineNumber++;
string line = *(++lr);
NStr::TruncateSpacesInPlace( line );
if (xProcessMetaLine(line, annot, pEC)) {
continue;
}
if (xProcessHeaderLine(line, annot)) {
continue;
}
if (xProcessDataLine(line, annot, pEC)) {
continue;
}
// still here? not good!
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Warning,
0,
"CVcfReader::ReadSeqAnnot: Unrecognized line or record type.",
ILineError::eProblem_GeneralParsingError) );
ProcessWarning(*pErr, pEC);
}
return annot;
}
// ---------------------------------------------------------------------------
bool
CVcfReader::xNormalizeData(
CVcfData& data,
IMessageListener* pEC)
// ---------------------------------------------------------------------------
{
// make sure none of the alternatives is equal to the reference:
for (size_t u=0; u < data.m_Alt.size(); ++u) {
if (data.m_Alt[u] == data.m_strRef) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
0,
"CVcfReader::xNormalizeData: Invalid alternative.",
ILineError::eProblem_GeneralParsingError));
ProcessError(*pErr, pEC);
return false;
}
}
// normalize ref/alt by trimming common prefices and adjusting location
bool trimComplete = false;
while (!data.m_strRef.empty()) {
char leadBase = data.m_strRef[0];
for (size_t u=0; u < data.m_Alt.size(); ++u) {
if (!NStr::StartsWith(data.m_Alt[u], leadBase)) {
trimComplete = true;
break;
}
}
if (trimComplete) {
break;
}
data.m_strRef = data.m_strRef.substr(1);
for (size_t u=0; u < data.m_Alt.size(); ++u) {
data.m_Alt[u] = data.m_Alt[u].substr(1);
}
data.m_iPos++;
}
// normalize ref/alt by trimming common postfixes and adjusting location
trimComplete = false;
size_t refSize = data.m_strRef.size();
size_t trimSize = 0;
while (refSize > trimSize) {
string postfix = data.m_strRef.substr(refSize-1-trimSize, trimSize+1);
for (size_t u=0; u < data.m_Alt.size(); ++u) {
size_t altSize = data.m_Alt[u].size();
if (altSize < trimSize+1) {
trimComplete = true;
break;
}
string postfixA = data.m_Alt[u].substr(altSize-1-trimSize, trimSize+1);
if (postfix != postfixA) {
trimComplete = true;
break;
}
}
if (trimComplete) {
break;
}
trimSize++;
}
if (trimSize > 0) {
data.m_strRef =
data.m_strRef.substr(0, data.m_strRef.size()-trimSize);
for (size_t u=0; u < data.m_Alt.size(); ++u) {
data.m_Alt[u] =
data.m_Alt[u].substr(0, data.m_Alt[u].size()-trimSize);
}
}
return true;
}
// ----------------------------------------------------------------------------
bool
CVcfReader::xParseData(
const string& line,
CVcfData& data,
IMessageListener* pEC)
// ----------------------------------------------------------------------------
{
vector<string> columns;
NStr::Tokenize( line, "\t", columns, NStr::eMergeDelims );
if ( columns.size() < 8 ) {
return false;
}
try {
data.m_strLine = line;
data.m_strChrom = columns[0];
data.m_iPos = NStr::StringToInt( columns[1] );
NStr::Tokenize( columns[2], ";", data.m_Ids, NStr::eNoMergeDelims );
if ( (data.m_Ids.size() == 1) && (data.m_Ids[0] == ".") ) {
data.m_Ids.clear();
}
data.m_strRef = columns[3];
NStr::Tokenize( columns[4], ",", data.m_Alt, NStr::eNoMergeDelims );
if ( columns[5] != "." ) {
data.m_pdQual = new double( NStr::StringToDouble( columns[5] ) );
}
data.m_strFilter = columns[6];
vector<string> infos;
if ( columns[7] != "." ) {
NStr::Tokenize( columns[7], ";", infos, NStr::eMergeDelims );
for ( vector<string>::iterator it = infos.begin();
it != infos.end(); ++it )
{
string key, value;
NStr::SplitInTwo( *it, "=", key, value );
data.m_Info[key] = vector<string>();
NStr::Tokenize( value, ",", data.m_Info[key] );
}
}
if ( columns.size() > 8 ) {
NStr::Tokenize( columns[8], ":", data.m_FormatKeys, NStr::eMergeDelims );
for ( size_t u=9; u < columns.size(); ++u ) {
vector<string> values;
NStr::Tokenize( columns[u], ":", values, NStr::eMergeDelims );
data.m_GenotypeData[ m_GenotypeHeaders[u-9] ] = values;
}
}
}
catch ( ... ) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
0,
"Unable to parse given VCF data (syntax error).",
ILineError::eProblem_GeneralParsingError));
ProcessError(*pErr, pEC);
return false;
}
if (!xNormalizeData(data, pEC)) {
return false;
}
//assign set type:
//test for all SNVs
bool maybeAllSnv = (data.m_strRef.size() == 1);
if (maybeAllSnv) {
for (size_t u=0; u < data.m_Alt.size(); ++u) {
if (data.m_Alt[u].size() != 1) {
maybeAllSnv = false;
break;
}
}
if (maybeAllSnv) {
data.m_SetType = CVcfData::ST_ALL_SNV;
return true;
}
}
//test for all mnvs:
bool maybeAllMnv = true;
size_t refSize = data.m_strRef.size();
for (size_t u=0; u < data.m_Alt.size(); ++u) {
if (data.m_Alt[u].size() != refSize) {
maybeAllMnv = false;
break;
}
}
if (maybeAllMnv) {
data.m_SetType = CVcfData::ST_ALL_MNV;
return true;
}
//test for all insertions:
bool maybeAllIns = true;
for (size_t u=0; u < data.m_Alt.size(); ++u) {
if (! NStr::StartsWith(data.m_Alt[u], data.m_strRef)) {
maybeAllIns = false;
break;
}
}
if (maybeAllIns) {
data.m_SetType = CVcfData::ST_ALL_INS;
return true;
}
//test for all deletions:
// note: even it is all deletions we are not able to process them
// as such because those deletions would be at different ASN1
// locations. Hence we punt to "indel" if there is more than one
// alternative.
bool maybeAllDel = false;
for (size_t u=0; u < data.m_Alt.size(); ++u) {
if (data.m_Alt.size() == 1 && data.m_Alt[0].empty()) {
maybeAllDel = true;
}
}
if (maybeAllDel) {
data.m_SetType = CVcfData::ST_ALL_DEL;
return true;
}
data.m_SetType = CVcfData::ST_MIXED;
return true;
}
// ----------------------------------------------------------------------------
bool
CVcfReader::xAssignVariantProps(
CVcfData& data,
CRef<CSeq_feat> pFeat,
IMessageListener* pEC)
// ----------------------------------------------------------------------------
{
typedef CVariantProperties VP;
CVcfData::INFOS& infos = data.m_Info;
VP& props = pFeat->SetData().SetVariation().SetVariant_prop();
CVcfData::INFOS::iterator it;
props.SetResource_link() = 0;
props.SetGene_location() = 0;
props.SetEffect() = 0;
props.SetMapping() = 0;
props.SetFrequency_based_validation() = 0;
props.SetGenotype() = 0;
props.SetQuality_check() = 0;
//byte F0
props.SetVersion() = 5;
//superbyte F1
it = infos.find("SLO");
if (infos.end() != it) {
props.SetResource_link() |= VP::eResource_link_submitterLinkout;
infos.erase(it);
}
it = infos.find("S3D");
if (infos.end() != it) {
props.SetResource_link() |= VP::eResource_link_has3D;
infos.erase(it);
}
it = infos.find("TPA");
if (infos.end() != it) {
props.SetResource_link() |= VP::eResource_link_provisional;
infos.erase(it);
}
it = infos.find("PM");
if (infos.end() != it) {
props.SetResource_link() |= VP::eResource_link_preserved;
infos.erase(it);
}
it = infos.find("CLN");
if (infos.end() != it) {
props.SetResource_link() |= VP::eResource_link_clinical;
infos.erase(it);
}
//todo: INFO ID=PMC
it = infos.find("PMC");
if (infos.end() != it) {
infos.erase(it);
}
it = infos.find("PMID");
if (infos.end() != it) {
vector<string> pmids = it->second;
for (vector<string>::const_iterator cit = pmids.begin();
cit != pmids.end(); ++cit)
{
try {
string db, tag;
NStr::SplitInTwo(*cit, ":", db, tag);
if (db != "PM") {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Warning,
0,
"CVcfReader::xAssignVariantProps: Invalid PMID database ID.",
ILineError::eProblem_GeneralParsingError) );
ProcessWarning(*pErr, pEC);
continue;
}
CRef<CDbtag> pDbtag(new CDbtag);
pDbtag->SetDb(db);
pDbtag->SetTag().SetId(
NStr::StringToInt(tag));
pFeat->SetDbxref().push_back(pDbtag);
}
catch(...) {}
}
infos.erase(it);
}
//superbyte F2
it = infos.find("R5");
if (infos.end() != it) {
props.SetGene_location() |= VP::eGene_location_near_gene_5;
infos.erase(it);
}
it = infos.find("R3");
if (infos.end() != it) {
props.SetGene_location() |= VP::eGene_location_near_gene_3;
infos.erase(it);
}
it = infos.find("INT");
if (infos.end() != it) {
props.SetGene_location() |= VP::eGene_location_intron;
infos.erase(it);
}
it = infos.find("DSS");
if (infos.end() != it) {
props.SetGene_location() |= VP::eGene_location_donor;
infos.erase(it);
}
it = infos.find("ASS");
if (infos.end() != it) {
props.SetGene_location() |= VP::eGene_location_acceptor;
infos.erase(it);
}
it = infos.find("U5");
if (infos.end() != it) {
props.SetGene_location() |= VP::eGene_location_utr_5;
infos.erase(it);
}
it = infos.find("U3");
if (infos.end() != it) {
props.SetGene_location() |= CVariantProperties::eGene_location_utr_3;
infos.erase(it);
}
it = infos.find("SYN");
if (infos.end() != it) {
props.SetGene_location() |= VP::eEffect_synonymous;
infos.erase(it);
}
it = infos.find("NSN");
if (infos.end() != it) {
props.SetGene_location() |= VP::eEffect_stop_gain;
infos.erase(it);
}
it = infos.find("NSM");
if (infos.end() != it) {
props.SetGene_location() |= VP::eEffect_missense;
infos.erase(it);
}
it = infos.find("NSF");
if (infos.end() != it) {
props.SetGene_location() |= VP::eEffect_frameshift;
infos.erase(it);
}
//byte F3
it = infos.find("WGT");
if (infos.end() != it) {
int weight = NStr::StringToInt( infos["WGT"][0] );
switch(weight) {
default:
break;
case 1:
props.SetMap_weight() = VP::eMap_weight_is_uniquely_placed;
infos.erase(it);
break;
case 2:
props.SetMap_weight() = VP::eMap_weight_placed_twice_on_same_chrom;
infos.erase(it);
break;
case 3:
props.SetMap_weight() = VP::eMap_weight_placed_twice_on_diff_chrom;
infos.erase(it);
break;
case 10:
props.SetMap_weight() = VP::eMap_weight_many_placements;
break;
}
}
it = infos.find("ASP");
if (infos.end() != it) {
props.SetMapping() |= VP::eMapping_is_assembly_specific;
infos.erase(it);
}
it = infos.find("CFL");
if (infos.end() != it) {
props.SetMapping() |= VP::eMapping_has_assembly_conflict;
infos.erase(it);
}
it = infos.find("OTH");
if (infos.end() != it) {
props.SetMapping() |= VP::eMapping_has_other_snp;
infos.erase(it);
}
//byte F4
it = infos.find("OTH");
if (infos.end() != it) {
props.SetMapping() |= VP::eFrequency_based_validation_above_5pct_all;
infos.erase(it);
}
it = infos.find("G5A");
if (infos.end() != it) {
props.SetMapping() |= VP::eFrequency_based_validation_above_5pct_1plus;
infos.erase(it);
}
it = infos.find("VLD");
if (infos.end() != it) {
props.SetMapping() |= VP::eFrequency_based_validation_validated;
infos.erase(it);
}
it = infos.find("MUT");
if (infos.end() != it) {
props.SetMapping() |= VP::eFrequency_based_validation_is_mutation;
infos.erase(it);
}
it = infos.find("GMAF");
if (infos.end() != it) {
props.SetAllele_frequency() = NStr::StringToDouble(infos["GMAF"][0]);
infos.erase(it);
}
//byte F5
it = infos.find("GNO");
if (infos.end() != it) {
props.SetGenotype() |= VP::eGenotype_has_genotypes;
infos.erase(it);
}
it = infos.find("HD");
if (infos.end() != it) {
props.SetResource_link() |= VP::eResource_link_genotypeKit;
infos.erase(it);
}
//byte F6
if (infos.end() != infos.find("PH3")) {
CRef<CDbtag> pDbtag(new CDbtag);
pDbtag->SetDb("BioProject");
pDbtag->SetTag().SetId(60835);
pFeat->SetData().SetVariation().SetOther_ids().push_back(pDbtag);
}
if (infos.end() != infos.find("KGPhase1")) {
CRef<CDbtag> pDbtag(new CDbtag);
pDbtag->SetDb("BioProject");
pDbtag->SetTag().SetId(28889);
pFeat->SetData().SetVariation().SetOther_ids().push_back(pDbtag);
}
//byte F7
//byte F8
//no relevant information found in VCF
//byte F9
it = infos.find("GCF");
if (infos.end() != it) {
props.SetQuality_check() |= VP::eQuality_check_genotype_conflict;
infos.erase(it);
}
it = infos.find("NOV");
if (infos.end() != it) {
props.SetQuality_check() |= VP::eQuality_check_non_overlapping_alleles;
infos.erase(it);
}
it = infos.find("WTD");
if (infos.end() != it) {
props.SetQuality_check() |= VP::eQuality_check_withdrawn_by_submitter;
infos.erase(it);
}
it = infos.find("NOC");
if (infos.end() != it) {
props.SetQuality_check() |= VP::eQuality_check_contig_allele_missing;
infos.erase(it);
}
return true;
}
// ----------------------------------------------------------------------------
void CUCSCRegionReader::x_SetFeatureLocation(
CRef<CSeq_feat>& feature,
const vector<string>& fields )
// ----------------------------------------------------------------------------
{
//
// Note:
// BED convention for specifying intervals is 0-based, first in, first out.
// ASN convention for specifying intervals is 0-based, first in, last in.
// Hence, conversion BED->ASN leaves the first leaves the "from" coordinate
// unchanged, and decrements the "to" coordinate by one.
//
CRef<CSeq_loc> location(new CSeq_loc);
int from, to;
from = to = -1;
//already established: We got at least three columns
try {
from = NStr::StringToInt(fields[1], NStr::fAllowCommas)-1;
}
catch(std::exception&) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
m_uLineNumber,
"Invalid data line: Bad \"SeqStart\" value." ) );
pErr->Throw();
}
to = from;
if (fields.size()>2)
try {
to = NStr::StringToInt(fields[2], NStr::fAllowCommas) - 1;
}
catch(std::exception&) {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
m_uLineNumber,
"Invalid data line: Bad \"SeqStop\" value.") );
pErr->Throw();
}
if (from == to) {
location->SetPnt().SetPoint(from);
}
else if (from < to) {
location->SetInt().SetFrom(from);
location->SetInt().SetTo(to);
}
else {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
m_uLineNumber,
"Invalid data line: \"SeqStop\" less than \"SeqStart\"." ) );
pErr->Throw();
}
size_t strand_field = 3;
if (strand_field < fields.size()) {
string strand = fields[strand_field];
if (strand != "+" && strand != "-" && strand != ".") {
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
m_uLineNumber,
"Invalid data line: Invalid strand character." ) );
pErr->Throw();
}
location->SetStrand(( fields[strand_field] == "+" ) ?
eNa_strand_plus : eNa_strand_minus );
}
try
{
CRef<CSeq_id> id = CReadUtil::AsSeqId(fields[0], m_iFlags, false);
//CRef<CSeq_id> id (new CSeq_id(fields[0], CSeq_id::fParse_AnyRaw | m_iFlags));
location->SetId(*id);
feature->SetLocation(*location);
}
catch(CSeqIdException&)
{
AutoPtr<CObjReaderLineException> pErr(
CObjReaderLineException::Create(
eDiag_Error,
m_uLineNumber,
"Malformed sequence id:" ) );
pErr->Throw();
}
}