// -----------------------------------------------------------------------------
bool CGtfReader::x_CreateParentMrna(
const CGff2Record& gff,
CRef< CSeq_annot > pAnnot )
// -----------------------------------------------------------------------------
{
//
// Create a single cds feature:
//
CRef< CSeq_feat > pFeature( new CSeq_feat );
if ( ! x_FeatureSetDataMRNA( gff, pFeature ) ) {
return false;
}
if ( ! x_CreateFeatureLocation( gff, pFeature ) ) {
return false;
}
if ( ! x_CreateFeatureId( gff, "mrna", pFeature ) ) {
return false;
}
if ( ! x_CreateGeneXrefs( gff, pFeature ) ) {
return false;
}
if ( ! x_FeatureSetQualifiers( gff, pFeature ) ) {
return false;
}
m_MrnaMap[ s_FeatureKey( gff ) ] = pFeature;
return xAddFeatureToAnnot( pFeature, pAnnot );
}
// -----------------------------------------------------------------------------
bool CGtfReader::x_CreateParentGene(
const CGff2Record& gff,
CRef< CSeq_annot > pAnnot )
// -----------------------------------------------------------------------------
{
//
// Create a single gene feature:
//
CRef< CSeq_feat > pFeature( new CSeq_feat );
if ( ! x_FeatureSetDataGene( gff, pFeature ) ) {
return false;
}
if ( ! x_CreateFeatureLocation( gff, pFeature ) ) {
return false;
}
if ( ! x_CreateFeatureId( gff, "gene", pFeature ) ) {
return false;
}
if ( ! xFeatureSetQualifiersGene( gff, pFeature ) ) {
return false;
}
m_GeneMap[ s_GeneKey( gff ) ] = pFeature;
xAddFeatureToAnnot( pFeature, pAnnot );
return true;
}
// ----------------------------------------------------------------------------
bool CGff3Reader::x_UpdateAnnot(
const CGff2Record& record,
CRef< CSeq_annot > pAnnot )
// ----------------------------------------------------------------------------
{
string gbkey;
record.GetAttribute("gbkey", gbkey);
CRef< CSeq_feat > pFeature(new CSeq_feat);
// Round trip info:
CRef< CGb_qual > pQual( new CGb_qual );
pQual->SetQual( "gff_source" );
pQual->SetVal( record.Source() );
pFeature->SetQual().push_back( pQual );
pQual.Reset( new CGb_qual );
pQual->SetQual( "gff_type" );
pQual->SetVal( record.Type() );
pFeature->SetQual().push_back( pQual );
if ( record.IsSetScore() ) {
pQual.Reset( new CGb_qual );
pQual->SetQual( "gff_score" );
pQual->SetVal( NStr::DoubleToString( record.Score() ) );
pFeature->SetQual().push_back( pQual );
}
// Special case: exon feature belonging to an RNA we have already seen
if (record.Type() == "exon") {
string parent;
if (record.GetAttribute("Parent", parent)) {
IdToFeatureMap::iterator it = m_MapIdToFeature.find(parent);
if (it != m_MapIdToFeature.end()) {
return record.UpdateFeature(m_iFlags, it->second);
}
}
}
// Special case: Piece of another feature we have already seen
string id;
if (record.GetAttribute("ID", id)) {
IdToFeatureMap::iterator it = m_MapIdToFeature.find(id);
if (it != m_MapIdToFeature.end()) {
return record.UpdateFeature(m_iFlags, it->second);
}
}
// General case: brand new regular feature
if (!record.InitializeFeature(m_iFlags, pFeature)) {
return false;
}
string strId;
if ( record.GetAttribute( "ID", strId ) ) {
m_MapIdToFeature[ strId ] = pFeature;
}
return x_AddFeatureToAnnot( pFeature, pAnnot );
}
// -----------------------------------------------------------------------------
bool CGtfReader::x_CreateParentCds(
const CGff2Record& gff,
CRef< CSeq_annot > pAnnot )
// -----------------------------------------------------------------------------
{
//
// Create a single cds feature.
// This creation may either be triggered by an actual CDS feature found in the
// gtf, or by a feature that would imply a CDS feature (such as a start codon
// or a stop codon). The latter is necessary because nothing the the gtf
// standard stipulates that gtf features have to be arranged in any particular
// order.
//
CRef< CSeq_feat > pFeature( new CSeq_feat );
string strType = gff.Type();
if ( strType != "CDS" && strType != "start_codon" && strType != "stop_codon" ) {
return false;
}
if ( ! x_FeatureSetDataCDS( gff, pFeature ) ) {
return false;
}
if ( ! x_CreateFeatureLocation( gff, pFeature ) ) {
return false;
}
if ( ! x_CreateFeatureId( gff, "cds", pFeature ) ) {
return false;
}
if ( ! x_CreateGeneXrefs( gff, pFeature ) ) {
return false;
}
if ( ! x_CreateMrnaXrefs( gff, pFeature ) ) {
return false;
}
if ( ! x_FeatureSetQualifiers( gff, pFeature ) ) {
return false;
}
m_CdsMap[ s_FeatureKey( gff ) ] = pFeature;
return xAddFeatureToAnnot( pFeature, pAnnot );
}
// ----------------------------------------------------------------------------
bool CGtfReader::x_UpdateAnnotFeature(
const CGff2Record& gff,
CRef< CSeq_annot > pAnnot,
ILineErrorListener* pEC)
// ----------------------------------------------------------------------------
{
CRef< CSeq_feat > pFeature( new CSeq_feat );
//
// Handle officially recognized GTF types:
//
string strType = gff.Type();
if ( strType == "CDS" ) {
//
// Observations:
// Location does not include the stop codon hence must be fixed up once
// the stop codon is seen.
//
return x_UpdateAnnotCds( gff, pAnnot );
}
if ( strType == "start_codon" ) {
//
// Observation:
// Comes in up to three pieces (depending on splicing).
// Location _is_ included in CDS.
//
return x_UpdateAnnotStartCodon( gff, pAnnot );
}
if ( strType == "stop_codon" ) {
//
// Observation:
// Comes in up to three pieces (depending on splicing).
// Location not included in CDS hence must be used to fix up location of
// the coding region.
//
return x_UpdateAnnotStopCodon( gff, pAnnot );
}
if ( strType == "5UTR" ) {
return x_UpdateAnnot5utr( gff, pAnnot );
}
if ( strType == "3UTR" ) {
return x_UpdateAnnot3utr( gff, pAnnot );
}
if ( strType == "inter" ) {
return x_UpdateAnnotInter( gff, pAnnot );
}
if ( strType == "inter_CNS" ) {
return x_UpdateAnnotInterCns( gff, pAnnot );
}
if ( strType == "intron_CNS" ) {
return x_UpdateAnnotIntronCns( gff, pAnnot );
}
if ( strType == "exon" ||
strType == "initial" ||
strType == "internal" ||
strType == "terminal" ||
strType == "single") {
return x_UpdateAnnotExon( gff, pAnnot );
}
//
// Every other type is not officially sanctioned GTF, and per spec we are
// supposed to ignore it. In the spirit of being lenient on input we may
// try to salvage some of it anyway.
//
if ( strType == "gene" ) {
//
// Not an official GTF feature type but seen frequently. Hence we give
// it some recognition.
//
if ( ! x_CreateParentGene( gff, pAnnot ) ) {
return false;
}
}
if (strType == "mRNA") {
if ( ! x_CreateParentMrna(gff, pAnnot) ) {
return false;
}
}
return x_UpdateAnnotMiscFeature( gff, pAnnot );
}
