bool buildUTGMessage(int32 ID, SnapUnitigMesg *utg) {
MultiAlignT *ma = ScaffoldGraph->tigStore->loadMultiAlign(ID, TRUE);
if (ma == NULL)
return(false);
utg->eaccession = AS_UID_fromInteger(getUID(uidServer));
utg->iaccession = ID;
utg->coverage_stat = ScaffoldGraph->tigStore->getUnitigCoverageStat(ID);
utg->microhet_prob = ScaffoldGraph->tigStore->getUnitigMicroHetProb(ID);
utg->status = ScaffoldGraph->tigStore->getUnitigStatus(ID);
utg->status = (utg->status == AS_UNASSIGNED ? AS_UNIQUE : utg->status);
utg->length = GetMultiAlignLength(ma);
utg->consensus = Getchar(ma->consensus, 0);
utg->quality = Getchar(ma->quality, 0);
utg->forced = 0;
utg->num_frags = GetNumIntMultiPoss(ma->f_list);
utg->num_vars = 0;
utg->f_list = (SnapMultiPos*)safe_malloc(utg->num_frags * sizeof(SnapMultiPos));
utg->v_list = NULL;
if (utg->consensus == NULL)
fprintf(stderr, "buildUTGMessage()-- unitig %d missing consensus sequence\n",
utg->iaccession);
assert(utg->consensus != NULL);
if (utg->length != strlen(utg->consensus))
fprintf(stderr, "buildUTGMessage()-- unitig %d length %d != consensus string length "F_SIZE_T"\n",
utg->iaccession, utg->length, strlen(utg->consensus));
assert(utg->length == strlen(utg->consensus));
for (int32 i=0; i<utg->num_frags; i++) {
IntMultiPos *imp = GetIntMultiPos(ma->f_list, i);
utg->f_list[i].type = imp->type;
utg->f_list[i].eident = FRGmap.lookup(imp->ident);
utg->f_list[i].position = imp->position;
utg->f_list[i].delta_length = imp->delta_length;
utg->f_list[i].delta = imp->delta;
}
return(true);
}
static
int
abAbacus::SetUngappedFragmentPositions(FragType type,int32 n_frags, MultiAlignT *uma) {
int32 num_frags = GetNumIntMultiPoss(uma->f_list);
int32 num_unitigs = GetNumIntUnitigPoss(uma->u_list);
HashTable_AS *unitigFrags = CreateScalarHashTable_AS();
int32 num_columns = GetMultiAlignLength(uma);
int32 ungapped_pos = 0;
int32 *gapped_positions = new int32 [num_columns + 1];
char *consensus = Getchar(uma->consensus,0);
for (int32 i=0; i<num_columns+1; i++) {
gapped_positions[i] = ungapped_pos;
if (consensus[i] != '-')
ungapped_pos++;
}
// Remember the first fragment we add.
int32 first_frag = GetNumCNS_AlignedContigElements(fragment_positions);
for (int32 ifrag=0; ifrag<num_frags; ifrag++) {
CNS_AlignedContigElement epos;
IntMultiPos *frag = GetIntMultiPos(uma->f_list, ifrag);
if (ExistsInHashTable_AS(unitigFrags, frag->ident, 0)) {
fprintf(stderr,"SetUngappedFragmentPositions()-- ident %d already in hashtable\n", frag->ident);
assert(0);
}
if (HASH_SUCCESS != InsertInHashTable_AS(unitigFrags, frag->ident, 0, 1, 0)) {
fprintf(stderr,"SetUngappedFragmentPositions()-- Failure to insert ident %d in hashtable\n", frag->ident);
assert(0);
}
assert(frag->position.bgn >= 0);
assert(frag->position.bgn < num_columns + 1);
assert(frag->position.end >= 0);
assert(frag->position.end < num_columns + 1);
epos.frg_or_utg = CNS_ELEMENT_IS_FRAGMENT;
epos.idx.fragment.frgIdent = frag->ident;
epos.idx.fragment.frgType = frag->type;
epos.idx.fragment.frgContained = frag->contained;
epos.idx.fragment.frgInUnitig = (type == AS_CONTIG) ? -1 : uma->maID;
epos.position.bgn = gapped_positions[frag->position.bgn];
epos.position.end = gapped_positions[frag->position.end];
//fprintf(stderr, "SetUngappedFragmentPositions()-- FRG id=%d type=%c pos=%d,%d (orig pos=%d,%d)\n",
// frag->ident, frag->type, epos.position.bgn, epos.position.end, frag->position.bgn, frag->position.end);
// Adjust the ungapped position if we fall within a gap
//
if (epos.position.bgn == epos.position.end) {
fprintf(stderr,"SetUngappedFragmentPositions()-- Encountered bgn==end=="F_S32" in ungapped coords within SetUngappedFragmentPositions for "F_CID "(gapped coords "F_S32","F_S32")\n",
epos.position.bgn,frag->ident,frag->position.bgn,frag->position.end);
assert(frag->position.bgn != frag->position.end);
if (frag->position.bgn < frag->position.end) {
if (epos.position.bgn > 0)
epos.position.bgn--;
else
epos.position.end++;
} else {
if (epos.position.end > 0)
epos.position.end--;
else
epos.position.bgn++;
}
fprintf(stderr,"SetUngappedFragmentPositions()-- Reset to "F_S32","F_S32"\n",
epos.position.bgn,
epos.position.end);
}
AppendVA_CNS_AlignedContigElement(fragment_positions, &epos);
}
for (int32 ifrag=0; ifrag < num_unitigs; ifrag++){
CNS_AlignedContigElement epos;
IntUnitigPos *unitig = GetIntUnitigPos(uma->u_list, ifrag);
epos.frg_or_utg = CNS_ELEMENT_IS_UNITIG;
epos.idx.unitig.utgIdent = unitig->ident;
epos.idx.unitig.utgType = unitig->type;
epos.position.bgn = gapped_positions[unitig->position.bgn];
epos.position.end = gapped_positions[unitig->position.end];
//fprintf(stderr, "SetUngappedFragmentPositions()-- UTG id=%d type=%c pos=%d,%d (orig pos=%d,%d)\n",
// unitig->ident, unitig->type, epos.position.bgn, epos.position.end, unitig->position.bgn, unitig->position.end);
AppendVA_CNS_AlignedContigElement(fragment_positions,&epos);
}
// This is used only by ReplaceEndUnitigInContig(). Mark fragments in the "anchoring" contig
// that belong to this unitig.
//
if (type != AS_CONTIG) {
Fragment *anchor = GetFragment(fragmentStore,0);
if ((anchor != NULL) &&
(anchor->type == AS_CONTIG)) {
CNS_AlignedContigElement *af = GetCNS_AlignedContigElement(fragment_positions, anchor->components);
for (int32 ifrag=0; ifrag < anchor->n_components; ifrag++, af++) {
if ((af->frg_or_utg == CNS_ELEMENT_IS_FRAGMENT) &&
(ExistsInHashTable_AS(unitigFrags, af->idx.fragment.frgIdent, 0)))
af->idx.fragment.frgInUnitig = uma->maID;
}
}
}
DeleteHashTable_AS(unitigFrags);
delete [] gapped_positions;
return first_frag;
}
void
writeCCO(FILE *asmFile, bool doWrite) {
SnapConConMesg cco;
GenericMesg pmesg = { &cco, MESG_CCO };
GraphNodeIterator contigs;
ContigT *contig;
fprintf(stderr, "writeCCO()--\n");
InitGraphNodeIterator(&contigs, ScaffoldGraph->ContigGraph, GRAPH_NODE_DEFAULT);
while ((contig = NextGraphNodeIterator(&contigs)) != NULL) {
assert(contig->id >= 0);
assert(contig->id < GetNumGraphNodes(ScaffoldGraph->ContigGraph));
if (contig->flags.bits.isChaff)
continue;
NodeCGW_T *unitig = GetGraphNode(ScaffoldGraph->CIGraph, contig->info.Contig.AEndCI);
if ((ScaffoldGraph->tigStore->getNumUnitigs(contig->id, FALSE) == 1) &&
(contig->scaffoldID == NULLINDEX) &&
(unitig->info.CI.numInstances > 0))
// Contig is a surrogate instance
continue;
MultiAlignT *ma = ScaffoldGraph->tigStore->loadMultiAlign(contig->id, FALSE);
cco.eaccession = AS_UID_fromInteger(getUID(uidServer));
cco.iaccession = contig->id;
cco.placed = ScaffoldGraph->tigStore->getContigStatus(contig->id);
cco.length = GetMultiAlignLength(ma);
cco.consensus = Getchar(ma->consensus, 0);
cco.quality = Getchar(ma->quality, 0);
cco.forced = 0;
cco.num_pieces = GetNumIntMultiPoss(ma->f_list);
cco.num_unitigs = GetNumIntMultiPoss(ma->u_list);
cco.num_vars = GetNumIntMultiPoss(ma->v_list);
cco.pieces = NULL;
cco.unitigs = NULL;
cco.vars = NULL;
if (cco.consensus == NULL)
fprintf(stderr, "buildCCOMessage()-- contig %d missing consensus sequence\n",
cco.iaccession);
assert(cco.consensus != NULL);
if (cco.length != strlen(cco.consensus))
fprintf(stderr, "buildCCOMessage()-- contig %d length %d != consensus string length "F_SIZE_T"\n",
cco.iaccession, cco.length, strlen(cco.consensus));
assert(cco.length == strlen(cco.consensus));
if (cco.num_pieces > 0) {
cco.pieces = (SnapMultiPos *)safe_malloc(cco.num_pieces * sizeof(SnapMultiPos));
for(int32 i=0; i<cco.num_pieces; i++) {
IntMultiPos *imp = GetIntMultiPos(ma->f_list, i);
cco.pieces[i].type = imp->type;
cco.pieces[i].eident = FRGmap.lookup(imp->ident);
cco.pieces[i].delta_length = imp->delta_length;
cco.pieces[i].position = imp->position;
cco.pieces[i].delta = imp->delta;
}
}
if (cco.num_unitigs > 0) {
cco.unitigs = (UnitigPos *)safe_malloc(cco.num_unitigs * sizeof(UnitigPos));
for(int32 i=0; i<cco.num_unitigs; i++) {
IntUnitigPos *imp = GetIntUnitigPos(ma->u_list, i);
cco.unitigs[i].type = imp->type;
cco.unitigs[i].eident = UTGmap.lookup(imp->ident);
cco.unitigs[i].position = imp->position;
cco.unitigs[i].delta = imp->delta;
cco.unitigs[i].delta_length = imp->delta_length;
}
}
if (cco.num_vars > 0) {
cco.vars = (IntMultiVar *)safe_malloc(cco.num_vars * sizeof(IntMultiVar));
for(int32 i=0; i<cco.num_vars; i++) {
IntMultiVar *imv = GetIntMultiVar(ma->v_list, i);
cco.vars[i].var_id = imv->var_id;
cco.vars[i].phased_id = imv->phased_id;
cco.vars[i].position = imv->position;
cco.vars[i].num_reads = imv->num_reads;
cco.vars[i].num_alleles = imv->num_alleles;
cco.vars[i].num_alleles_confirmed = imv->num_alleles_confirmed;
cco.vars[i].min_anchor_size = imv->min_anchor_size;
cco.vars[i].var_length = imv->var_length;
cco.vars[i].alleles = imv->alleles;
cco.vars[i].var_seq_memory = imv->var_seq_memory;
cco.vars[i].read_id_memory = imv->read_id_memory;
cco.vars[i].enc_num_reads = NULL;
cco.vars[i].enc_weights = NULL;
cco.vars[i].enc_var_seq = NULL;
cco.vars[i].enc_read_ids = NULL;
}
}
if (doWrite)
WriteProtoMesg_AS(asmFile, &pmesg);
safe_free(cco.pieces);
safe_free(cco.unitigs);
safe_free(cco.vars);
CCOmap.add(cco.iaccession, cco.eaccession);
}
}