static void split_cds_feature(GtFeatureNode *cds_feature, GtFeatureNode *fn)
{
GtArray *parents;
unsigned long i;
gt_assert(cds_feature && fn);
/* find parents */
parents = find_cds_parents(cds_feature, fn);
/* remove CDS feature */
gt_feature_node_remove_leaf(fn, cds_feature);
/* add CDS feature to all parents */
for (i = 0; i < gt_array_size(parents); i++) {
GtFeatureNode *parent = *(GtFeatureNode**) gt_array_get(parents, i);
const char *id = gt_feature_node_get_attribute(parent, GT_GFF_ID);
if (!i) {
gt_feature_node_set_attribute(cds_feature, GT_GFF_PARENT, id);
gt_feature_node_add_child(parent, cds_feature);
}
else {
GtFeatureNode *new_cds = gt_feature_node_clone(cds_feature);
gt_feature_node_set_attribute(new_cds, GT_GFF_PARENT, id);
gt_feature_node_add_child(parent, new_cds);
gt_genome_node_delete((GtGenomeNode*) cds_feature);
}
}
gt_array_delete(parents);
}
static int pdom_hit_attach_gff3(GtPdomModel *model, GtPdomModelHit *hit,
void *data, GT_UNUSED GtError *err)
{
unsigned long i;
GtRange rng;
GtLTRdigestStream *ls = (GtLTRdigestStream *) data;
GtStrand strand;
gt_assert(model && hit);
strand = gt_pdom_model_hit_get_best_strand(hit);
/* do not use the hits on the non-predicted strand
-- maybe identify nested elements ? */
if (strand != gt_feature_node_get_strand(ls->element.mainnode))
return 0;
for (i=0;i<gt_pdom_model_hit_best_chain_length(hit);i++)
{
GtGenomeNode *gf;
GtStr *alignmentstring,
*aastring;
GtPdomSingleHit *singlehit;
GtPhase frame;
singlehit = gt_pdom_model_hit_best_single_hit(hit, i);
alignmentstring = gt_str_new();
aastring = gt_str_new();
frame = gt_pdom_single_hit_get_phase(singlehit);
rng = gt_pdom_single_hit_get_range(singlehit);
gt_pdom_single_hit_format_alignment(singlehit, GT_ALIWIDTH,
alignmentstring);
gt_pdom_single_hit_get_aaseq(singlehit, aastring);
rng.start++; rng.end++; /* GFF3 is 1-based */
gf = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*)
ls->element.mainnode),
GT_PDOM_TYPE,
rng.start,
rng.end,
strand);
gt_genome_node_add_user_data((GtGenomeNode*) gf, "pdom_alignment",
alignmentstring, (GtFree) gt_str_delete);
gt_genome_node_add_user_data((GtGenomeNode*) gf, "pdom_aaseq",
aastring, (GtFree) gt_str_delete);
gt_feature_node_set_source((GtFeatureNode*) gf, ls->ltrdigest_tag);
gt_feature_node_set_score((GtFeatureNode*) gf,
gt_pdom_single_hit_get_evalue(singlehit));
gt_feature_node_set_phase((GtFeatureNode*) gf, frame);
if (gt_pdom_model_get_name(model)) {
gt_feature_node_add_attribute((GtFeatureNode*) gf, "name",
gt_pdom_model_get_name(model));
}
if (gt_pdom_model_get_acc(model)) {
gt_feature_node_add_attribute((GtFeatureNode*) gf, "id",
gt_pdom_model_get_acc(model));
}
gt_feature_node_add_child(ls->element.mainnode, (GtFeatureNode*) gf);
}
return 0;
}
static int gt_ltrdigest_pdom_visitor_attach_hit(GtLTRdigestPdomVisitor *lv,
GtHMMERModelHit *modelhit,
GtHMMERSingleHit *singlehit)
{
GT_UNUSED GtUword i;
GtGenomeNode *gf;
int had_err = 0;
GtRange rrng;
gt_assert(lv && singlehit);
rrng = gt_ltrdigest_pdom_visitor_coords(lv, singlehit);
if (gt_array_size(singlehit->chains) > 0 || lv->output_all_chains) {
char buf[32];
gf = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*)
lv->ltr_retrotrans),
gt_ft_protein_match,
rrng.start,
rrng.end,
singlehit->strand);
gt_genome_node_add_user_data((GtGenomeNode*) gf, "pdom_alignment",
gt_str_ref(singlehit->alignment),
(GtFree) gt_str_delete);
gt_genome_node_add_user_data((GtGenomeNode*) gf, "pdom_aaseq",
gt_str_ref(singlehit->aastring),
(GtFree) gt_str_delete);
gt_feature_node_set_source((GtFeatureNode*) gf, lv->tag);
gt_feature_node_set_score((GtFeatureNode*) gf, (float) singlehit->evalue);
(void) snprintf(buf, (size_t) 32, "%d", (int) singlehit->frame);
gt_feature_node_add_attribute((GtFeatureNode*) gf,
"reading_frame", buf);
if (modelhit->modelname != NULL) {
gt_feature_node_add_attribute((GtFeatureNode*) gf, "name",
modelhit->modelname);
}
if (gt_array_size(singlehit->chains) > 1UL && lv->output_all_chains) {
GtStr *buffer;
GtUword j;
gt_assert(singlehit->chains != NULL);
buffer = gt_str_new();
for (j = 0UL; j < gt_array_size(singlehit->chains); j++) {
gt_str_append_cstr(buffer, modelhit->modelname);
gt_str_append_char(buffer, ':');
gt_str_append_ulong(buffer,
*(GtUword*) gt_array_get(singlehit->chains, j));
if (j != gt_array_size(singlehit->chains) - 1) {
gt_str_append_char(buffer, ',');
}
}
gt_feature_node_set_attribute((GtFeatureNode*) gf, "chains",
gt_str_get(buffer));
gt_str_delete(buffer);
}
gt_feature_node_add_child(lv->ltr_retrotrans, (GtFeatureNode*) gf);
}
gt_array_delete(singlehit->chains);
singlehit->chains = NULL;
return had_err;
}
static void infer_cds_visitor_check_stop(AgnInferCDSVisitor *v)
{
if(gt_array_size(v->cds) == 0)
return;
const char *mrnaid = gt_feature_node_get_attribute(v->mrna, "ID");
unsigned int ln = gt_genome_node_get_line_number((GtGenomeNode *)v->mrna);
GtStrand strand = gt_feature_node_get_strand(v->mrna);
GtRange stoprange;
GtUword threeprimeindex = gt_array_size(v->cds) - 1;
GtGenomeNode **threeprimesegment = gt_array_get(v->cds, threeprimeindex);
stoprange = gt_genome_node_get_range(*threeprimesegment);
stoprange.start = stoprange.end - 2;
if(strand == GT_STRAND_REVERSE)
{
threeprimesegment = gt_array_get(v->cds, 0);
stoprange = gt_genome_node_get_range(*threeprimesegment);
stoprange.end = stoprange.start + 2;
}
if(gt_array_size(v->stops) > 1)
{
gt_logger_log(v->logger, "mRNA '%s' (line %u) has %lu stop codons", mrnaid,
ln, gt_array_size(v->starts));
}
else if(gt_array_size(v->stops) == 1)
{
GtGenomeNode **codon = gt_array_get(v->stops, 0);
GtRange testrange = gt_genome_node_get_range(*codon);
if(gt_range_compare(&stoprange, &testrange) != 0)
{
gt_logger_log(v->logger, "stop codon inferred from CDS [%lu, %lu] does "
"not match explicitly provided stop codon [%lu, %lu] for "
"mRNA '%s'", stoprange.start, stoprange.end,
testrange.start, testrange.end, mrnaid);
}
}
else // agn_assert(gt_array_size(v->stops) == 0)
{
GtStr *seqid = gt_genome_node_get_seqid((GtGenomeNode *)v->mrna);
GtGenomeNode *codonfeature = gt_feature_node_new(seqid, "stop_codon",
stoprange.start,
stoprange.end,
strand);
if(v->source)
gt_feature_node_set_source((GtFeatureNode *)codonfeature, v->source);
GtFeatureNode *cf = (GtFeatureNode *)codonfeature;
gt_feature_node_add_child(v->mrna, cf);
gt_array_add(v->stops, cf);
}
}
static int create_block_features(GtBEDParser *bed_parser, GtFeatureNode *fn,
GtUword block_count,
GtSplitter *size_splitter,
GtSplitter *start_splitter, GtIO *bed_file,
GtError *err)
{
GtUword i;
int had_err = 0;
gt_assert(fn && block_count && size_splitter && start_splitter);
gt_assert(gt_splitter_size(size_splitter) == block_count);
gt_assert(gt_splitter_size(start_splitter) == block_count);
for (i = 0; !had_err && i < block_count; i++) {
GtUword block_size, block_start, start, end;
GtGenomeNode *block;
const char *name;
if (gt_parse_uword(&block_size, gt_splitter_get_token(size_splitter, i))) {
gt_error_set(err,
"file \"%s\": line "GT_WU": could not parse blockSize '%s'",
gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file),
gt_splitter_get_token(size_splitter, i));
had_err = -1;
}
if (!had_err && gt_parse_uword(&block_start,
gt_splitter_get_token(start_splitter, i))) {
gt_error_set(err, "file \"%s\": line "GT_WU": could not parse blockStart "
"'%s'", gt_io_get_filename(bed_file),
gt_io_get_line_number(bed_file),
gt_splitter_get_token(start_splitter, i));
had_err = -1;
}
if (!had_err) {
start = gt_genome_node_get_start((GtGenomeNode*) fn) + block_start;
end = start + block_size - 1;
block = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*) fn),
bed_parser->block_type
? bed_parser->block_type
: BED_BLOCK_TYPE,
start, end, gt_feature_node_get_strand(fn));
if ((name = gt_feature_node_get_attribute(fn, GT_GFF_NAME))) {
gt_feature_node_add_attribute((GtFeatureNode*) block, GT_GFF_NAME,
name);
}
gt_feature_node_set_score((GtFeatureNode*) block,
gt_feature_node_get_score(fn));
gt_feature_node_set_strand((GtFeatureNode*) block,
gt_feature_node_get_strand(fn));
gt_feature_node_add_child(fn, (GtFeatureNode*) block);
}
}
return had_err;
}
static void infer_cds_visitor_infer_cds(AgnInferCDSVisitor *v)
{
GtFeatureNode **start_codon = NULL, **stop_codon = NULL;
bool exonsexplicit = gt_array_size(v->exons) > 0;
bool startcodon_check = gt_array_size(v->starts) == 1 &&
(start_codon = gt_array_get(v->starts, 0)) != NULL;
bool stopcodon_check = gt_array_size(v->stops) == 1 &&
(stop_codon = gt_array_get(v->stops, 0)) != NULL;
if(gt_array_size(v->cds) > 0)
{
return;
}
else if(!exonsexplicit || !startcodon_check || !stopcodon_check)
{
return;
}
GtRange left_codon_range, right_codon_range;
left_codon_range = gt_genome_node_get_range(*(GtGenomeNode **)start_codon);
right_codon_range = gt_genome_node_get_range(*(GtGenomeNode **)stop_codon);
if(gt_feature_node_get_strand(v->mrna) == GT_STRAND_REVERSE)
{
left_codon_range = gt_genome_node_get_range(*(GtGenomeNode **)stop_codon);
right_codon_range = gt_genome_node_get_range(*(GtGenomeNode **)start_codon);
}
GtUword i;
for(i = 0; i < gt_array_size(v->exons); i++)
{
GtFeatureNode *exon = *(GtFeatureNode **)gt_array_get(v->exons, i);
GtGenomeNode *exon_gn = (GtGenomeNode *)exon;
GtRange exon_range = gt_genome_node_get_range(exon_gn);
GtStrand exon_strand = gt_feature_node_get_strand(exon);
GtRange cdsrange;
bool exon_includes_cds = infer_cds_visitor_infer_range(&exon_range,
&left_codon_range,
&right_codon_range,
&cdsrange);
if(exon_includes_cds)
{
GtGenomeNode *cdsfeat;
cdsfeat = gt_feature_node_new(gt_genome_node_get_seqid(exon_gn), "CDS",
cdsrange.start, cdsrange.end, exon_strand);
if(v->source)
gt_feature_node_set_source((GtFeatureNode *)cdsfeat, v->source);
gt_feature_node_add_child(v->mrna, (GtFeatureNode *)cdsfeat);
gt_array_add(v->cds, cdsfeat);
}
}
}
static void pbs_attach_results_to_gff3(GtPBSResults *results,
GtLTRElement *element,
GtStrand *canonical_strand,
GtStr *tag)
{
GtRange pbs_range;
GtGenomeNode *gf;
unsigned long i = 0;
char buffer[BUFSIZ];
GtPBSHit* hit = gt_pbs_results_get_ranked_hit(results, i++);
if (*canonical_strand == GT_STRAND_UNKNOWN)
*canonical_strand = gt_pbs_hit_get_strand(hit);
else
{
/* do we have to satisfy a strand constraint?
* then find best-scoring PBS on the given canonical strand */
while (gt_pbs_hit_get_strand(hit) != *canonical_strand
&& i < gt_pbs_results_get_number_of_hits(results))
{
gt_log_log("dropping PBS because of nonconsistent strand: %s\n",
gt_feature_node_get_attribute(element->mainnode, "ID"));
hit = gt_pbs_results_get_ranked_hit(results, i++);
}
/* if there is none, do not report a PBS */
if (gt_pbs_hit_get_strand(hit) != *canonical_strand)
return;
}
pbs_range = gt_pbs_hit_get_coords(hit);
pbs_range.start++; pbs_range.end++; /* GFF3 is 1-based */
gf = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*)
element->mainnode),
GT_PBS_TYPE,
pbs_range.start,
pbs_range.end,
gt_pbs_hit_get_strand(hit));
gt_feature_node_set_source((GtFeatureNode*) gf, tag);
gt_feature_node_set_score((GtFeatureNode*) gf,
(float) gt_pbs_hit_get_score(hit));
if (gt_pbs_hit_get_trna(hit) != NULL) {
gt_feature_node_add_attribute((GtFeatureNode*) gf, "trna",
gt_pbs_hit_get_trna(hit));
}
gt_feature_node_set_strand(element->mainnode, gt_pbs_hit_get_strand(hit));
(void) snprintf(buffer, BUFSIZ-1, "%lu", gt_pbs_hit_get_tstart(hit));
gt_feature_node_add_attribute((GtFeatureNode*) gf, "trnaoffset", buffer);
(void) snprintf(buffer, BUFSIZ-1, "%lu", gt_pbs_hit_get_offset(hit));
gt_feature_node_add_attribute((GtFeatureNode*) gf, "pbsoffset", buffer);
(void) snprintf(buffer, BUFSIZ-1, "%lu", gt_pbs_hit_get_edist(hit));
gt_feature_node_add_attribute((GtFeatureNode*) gf, "edist", buffer);
gt_feature_node_add_child(element->mainnode, (GtFeatureNode*) gf);
}
static int genome_node_lua_add_child(lua_State *L)
{
GtGenomeNode **parent, **child;
GtFeatureNode *pf, *cf;
parent = check_genome_node(L, 1);
child = check_genome_node(L, 2);
pf = gt_feature_node_try_cast(*parent);
luaL_argcheck(L, pf, 1, "not a feature node");
cf = gt_feature_node_try_cast(*child);
luaL_argcheck(L, cf, 2, "not a feature node");
gt_feature_node_add_child(pf, (GtFeatureNode*)
gt_genome_node_ref((GtGenomeNode*) cf));
return 0;
}
static void construct_thick_feature(GtBEDParser *bed_parser, GtFeatureNode *fn,
GtRange range)
{
GtGenomeNode *thick_feature;
const char *name;
gt_assert(fn);
thick_feature = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*)
fn),
bed_parser->thick_feature_type
? bed_parser->thick_feature_type
: BED_THICK_FEATURE_TYPE,
range.start, range.end,
gt_feature_node_get_strand(fn));
if ((name = gt_feature_node_get_attribute(fn, "Name")))
gt_feature_node_add_attribute((GtFeatureNode*) thick_feature, "Name", name);
gt_feature_node_set_score((GtFeatureNode*) thick_feature,
gt_feature_node_get_score(fn));
gt_feature_node_set_strand((GtFeatureNode*) thick_feature,
gt_feature_node_get_strand(fn));
gt_feature_node_add_child(fn, (GtFeatureNode*) thick_feature);
}
static void orf_attach_results_to_gff3(GtFeatureNode *gf,
GtRange orf_rng, unsigned int orf_frame,
GtStrand strand, GT_UNUSED GtError *err)
{
GtGenomeNode *child;
GtStr *tag;
tag = gt_str_new_cstr(GT_ORF_FINDER_TAG);
orf_rng.start++; orf_rng.end++;
GtFeatureNodeIterator *gfi;
GtFeatureNode *curnode = NULL, *parent_node = NULL;
GtRange gfi_range;
char frame_buf[3];
sprintf(frame_buf, "%d", orf_frame);
gfi = gt_feature_node_iterator_new(gf);
while ((curnode = gt_feature_node_iterator_next(gfi))) {
if (strcmp(gt_feature_node_get_type(curnode),
(const char*) GT_ORF_TYPE) != 0) {
gfi_range = gt_genome_node_get_range((GtGenomeNode*) curnode);
if (gt_range_contains(&gfi_range, &orf_rng)) {
parent_node = curnode;
}
}
}
if (parent_node) {
child = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*) gf),
GT_ORF_TYPE,
orf_rng.start,
orf_rng.end,
strand);
gt_feature_node_set_source((GtFeatureNode*) child, tag);
gt_feature_node_set_attribute((GtFeatureNode*) child, "frame", frame_buf);
gt_feature_node_add_child(parent_node,(GtFeatureNode*) child);
}
gt_str_delete(tag);
gt_feature_node_iterator_delete(gfi);
}
static void ppt_attach_results_to_gff3(GtPPTResults *results,
GtLTRElement *element,
GtStrand *canonical_strand,
GtStr *tag)
{
GtRange ppt_range;
unsigned long i = 0;
GtGenomeNode *gf;
GtPPTHit* hit = gt_ppt_results_get_ranked_hit(results, i++);
if (*canonical_strand == GT_STRAND_UNKNOWN)
*canonical_strand = gt_ppt_hit_get_strand(hit);
else
{
/* find best-scoring PPT on the given canonical strand */
while (gt_ppt_hit_get_strand(hit) != *canonical_strand
&& i < gt_ppt_results_get_number_of_hits(results))
{
gt_log_log("dropping PPT because of nonconsistent strand: %s\n",
gt_feature_node_get_attribute(element->mainnode, "ID"));
hit = gt_ppt_results_get_ranked_hit(results, i++);
}
/* if there is none, do not report a PPT */
if (gt_ppt_hit_get_strand(hit) != *canonical_strand)
return;
}
ppt_range = gt_ppt_hit_get_coords(hit);
ppt_range.start++; ppt_range.end++; /* GFF3 is 1-based */
gf = gt_feature_node_new(gt_genome_node_get_seqid((GtGenomeNode*)
element->mainnode),
GT_PPT_TYPE,
ppt_range.start,
ppt_range.end,
gt_ppt_hit_get_strand(hit));
gt_feature_node_set_source((GtFeatureNode*) gf, tag);
gt_feature_node_set_strand(element->mainnode, gt_ppt_hit_get_strand(hit));
gt_feature_node_add_child(element->mainnode, (GtFeatureNode*) gf);
}
static int inter_feature_in_children(GtFeatureNode *current_feature, void *data,
GT_UNUSED GtError *err)
{
GtInterFeatureVisitor *aiv = (GtInterFeatureVisitor*) data;
GtFeatureNode *inter_node;
GtRange previous_range, current_range, inter_range;
GtStrand previous_strand, /*current_strand, */inter_strand;
GtStr *parent_seqid;
gt_error_check(err);
gt_assert(current_feature);
if (gt_feature_node_has_type(current_feature, aiv->outside_type)) {
if (aiv->previous_feature) {
/* determine inter range */
previous_range = gt_genome_node_get_range((GtGenomeNode*)
aiv->previous_feature);
current_range = gt_genome_node_get_range((GtGenomeNode*) current_feature);
if (previous_range.end >= current_range.start) {
gt_warning("overlapping boundary features " GT_WU "-" GT_WU " and "
GT_WU "-" GT_WU ", " "not placing '%s' inter-feature",
previous_range.start,
previous_range.end,
current_range.start,
current_range.end,
aiv->inter_type);
return 0;
}
if (current_range.start - previous_range.end < 2) {
gt_warning("no space for inter-feature '%s' between " GT_WU " and "
GT_WU,
aiv->inter_type,
previous_range.end,
current_range.start);
return 0;
}
inter_range.start = previous_range.end + 1;
inter_range.end = current_range.start - 1;
/* determine inter strand */
previous_strand = gt_feature_node_get_strand(aiv->previous_feature);
/*current_strand = gt_feature_node_get_strand(current_feature);*/
gt_assert(previous_strand == gt_feature_node_get_strand(current_feature));
inter_strand = previous_strand;
/* determine sequence id */
parent_seqid =
gt_genome_node_get_seqid((GtGenomeNode*) aiv->parent_feature);
gt_assert(!gt_str_cmp(parent_seqid,
gt_genome_node_get_seqid((GtGenomeNode*)
aiv->previous_feature)));
gt_assert(!gt_str_cmp(parent_seqid,
gt_genome_node_get_seqid((GtGenomeNode*)
current_feature)));
/* create inter feature */
inter_node = (GtFeatureNode*)
gt_feature_node_new(parent_seqid, aiv->inter_type,
inter_range.start, inter_range.end,
inter_strand);
gt_feature_node_add_child(aiv->parent_feature, inter_node);
}
aiv->previous_feature = current_feature;
}
return 0;
}
static int construct_mRNAs(GT_UNUSED void *key, void *value, void *data,
GtError *err)
{
ConstructionInfo *cinfo = (ConstructionInfo*) data;
GtArray *gt_genome_node_array = (GtArray*) value,
*mRNAs = (GtArray*) cinfo->mRNAs;
GtGenomeNode *mRNA_node, *first_node, *gn;
const char *tname;
GtStrand mRNA_strand;
GtRange mRNA_range;
GtStr *mRNA_seqid;
GtUword i;
int had_err = 0;
gt_error_check(err);
gt_assert(key && value && data);
/* at least one node in array */
gt_assert(gt_array_size(gt_genome_node_array));
/* determine the range and the strand of the mRNA */
first_node = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, 0);
mRNA_range = gt_genome_node_get_range(first_node);
mRNA_strand = gt_feature_node_get_strand((GtFeatureNode*) first_node);
mRNA_seqid = gt_genome_node_get_seqid(first_node);
/* TODO: support discontinuous start/stop codons */
for (i = 0; !had_err && i < gt_array_size(gt_genome_node_array); i++) {
gn = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, i);
if (gt_feature_node_get_attribute((GtFeatureNode*) gn,
GTF_PARSER_STOP_CODON_FLAG)) {
GtUword j;
GtRange stop_codon_rng = gt_genome_node_get_range(gn);
bool found_cds = false;
for (j = 0; !had_err && j < gt_array_size(gt_genome_node_array); j++) {
GtGenomeNode* gn2;
GtRange this_rng;
const char *this_type;
gn2 = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, j);
if (gn == gn2) continue;
this_rng = gt_genome_node_get_range(gn2);
this_type = gt_feature_node_get_type((GtFeatureNode*) gn2);
if (this_type == gt_symbol(gt_ft_CDS)) {
if (gt_range_contains(&this_rng, &stop_codon_rng)) {
if (cinfo->tidy) {
gt_warning("stop codon on line %u in file %s is contained in "
"CDS in line %u",
gt_genome_node_get_line_number(gn),
gt_genome_node_get_filename(gn),
gt_genome_node_get_line_number(gn2));
found_cds = true;
} else {
gt_error_set(err, "stop codon on line %u in file %s is "
"contained in CDS in line %u",
gt_genome_node_get_line_number(gn),
gt_genome_node_get_filename(gn),
gt_genome_node_get_line_number(gn2));
had_err = -1;
}
break;
}
if (this_rng.end + 1 == stop_codon_rng.start) {
this_rng.end = stop_codon_rng.end;
gt_genome_node_set_range(gn2, &this_rng);
found_cds = true;
break;
}
if (this_rng.start == stop_codon_rng.end + 1) {
this_rng.start = stop_codon_rng.start;
gt_genome_node_set_range(gn2, &this_rng);
found_cds = true;
break;
}
}
}
if (!found_cds) {
if (!had_err) {
if (cinfo->tidy) {
gt_warning("found stop codon on line %u in file %s with no "
"flanking CDS, ignoring it",
gt_genome_node_get_line_number(gn),
gt_genome_node_get_filename(gn));
} else {
gt_error_set(err, "found stop codon on line %u in file %s with no "
"flanking CDS",
gt_genome_node_get_line_number(gn),
gt_genome_node_get_filename(gn));
had_err = -1;
break;
}
}
} else {
gt_array_rem(gt_genome_node_array, i);
gt_genome_node_delete(gn);
}
}
}
for (i = 1; !had_err && i < gt_array_size(gt_genome_node_array); i++) {
GtRange range;
GtStrand strand;
gn = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, i);
range = gt_genome_node_get_range(gn);
mRNA_range = gt_range_join(&mRNA_range, &range);
strand = gt_feature_node_get_strand((GtFeatureNode*) gn);
if (strand != mRNA_strand) {
gt_error_set(err, "feature %s on line %u has strand %c, but the "
"parent transcript has strand %c",
(const char*) key,
gt_genome_node_get_line_number(gn),
GT_STRAND_CHARS[strand],
GT_STRAND_CHARS[mRNA_strand]);
had_err = -1;
break;
} else {
mRNA_strand = gt_strand_join(mRNA_strand, strand);
}
if (!had_err && gt_str_cmp(mRNA_seqid, gt_genome_node_get_seqid(gn))) {
gt_error_set(err, "The features on lines %u and %u refer to different "
"genomic sequences (``seqname''), although they have the same "
"gene IDs (``gene_id'') which must be globally unique",
gt_genome_node_get_line_number(first_node),
gt_genome_node_get_line_number(gn));
had_err = -1;
break;
}
}
if (!had_err) {
mRNA_node = gt_feature_node_new(mRNA_seqid, gt_ft_mRNA, mRNA_range.start,
mRNA_range.end, mRNA_strand);
gt_feature_node_add_attribute(((GtFeatureNode*) mRNA_node), "ID", key);
gt_feature_node_add_attribute(((GtFeatureNode*) mRNA_node), "transcript_id",
key);
if ((tname = gt_hashmap_get(cinfo->transcript_id_to_name_mapping,
(const char*) key)) && strlen(tname) > 0) {
gt_feature_node_add_attribute((GtFeatureNode*) mRNA_node, GT_GFF_NAME,
tname);
}
/* register children */
for (i = 0; i < gt_array_size(gt_genome_node_array); i++) {
gn = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, i);
gt_feature_node_add_child((GtFeatureNode*) mRNA_node,
(GtFeatureNode*) gt_genome_node_ref(gn));
}
/* store the mRNA */
gt_array_add(mRNAs, mRNA_node);
}
return had_err;
}
GtGenomeNode* gt_feature_node_new_standard_gene(void)
{
GtGenomeNode *fn, *child, *grand;
GtStr *seqid;
seqid = gt_str_new_cstr("ctg123");
/* gene */
fn = gt_feature_node_new(seqid, gt_ft_gene, 1000, 9000, GT_STRAND_FORWARD);
/* TF binding site */
child = gt_feature_node_new(seqid, gt_ft_TF_binding_site, 1000, 1012,
GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) fn, (GtFeatureNode*) child);
/* first mRNA */
child = gt_feature_node_new(seqid, gt_ft_mRNA, 1050, 9000, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) fn, (GtFeatureNode*) child);
grand = gt_feature_node_new(seqid, gt_ft_exon, 1050, 1500, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 3000, 3902, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 5000, 5500, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 7000, 9000, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
/* second mRNA */
child = gt_feature_node_new(seqid, gt_ft_mRNA, 1050, 9000, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) fn, (GtFeatureNode*) child);
grand = gt_feature_node_new(seqid, gt_ft_exon, 1050, 1500, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 5000, 5500, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 7000, 9000, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
/* third mRNA */
child = gt_feature_node_new(seqid, gt_ft_mRNA, 1300, 9000, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) fn, (GtFeatureNode*) child);
grand = gt_feature_node_new(seqid, gt_ft_exon, 1300, 1500, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 3000, 3902, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 5000, 5500, GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
grand = gt_feature_node_new(seqid, gt_ft_exon, 7000, 9000,GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) child, (GtFeatureNode*) grand);
gt_str_delete(seqid);
return fn;
}
static int construct_mRNAs(GT_UNUSED void *key, void *value, void *data,
GtError *err)
{
ConstructionInfo *cinfo = (ConstructionInfo*) data;
GtArray *gt_genome_node_array = (GtArray*) value,
*mRNAs = (GtArray*) cinfo->mRNAs;
GtGenomeNode *mRNA_node, *first_node, *gn;
const char *tname;
GtStrand mRNA_strand;
GtRange mRNA_range;
GtStr *mRNA_seqid;
GtUword i;
int had_err = 0;
gt_error_check(err);
gt_assert(key && value && data);
/* at least one node in array */
gt_assert(gt_array_size(gt_genome_node_array));
/* determine the range and the strand of the mRNA */
first_node = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, 0);
mRNA_range = gt_genome_node_get_range(first_node);
mRNA_strand = gt_feature_node_get_strand((GtFeatureNode*) first_node);
mRNA_seqid = gt_genome_node_get_seqid(first_node);
for (i = 1; i < gt_array_size(gt_genome_node_array); i++) {
GtRange range;
gn = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, i);
range = gt_genome_node_get_range(gn);
mRNA_range = gt_range_join(&mRNA_range, &range);
/* XXX: an error check is necessary here, otherwise gt_strand_join() can
cause a failed assertion */
mRNA_strand = gt_strand_join(mRNA_strand,
gt_feature_node_get_strand((GtFeatureNode*) gn));
if (gt_str_cmp(mRNA_seqid, gt_genome_node_get_seqid(gn))) {
gt_error_set(err, "The features on lines %u and %u refer to different "
"genomic sequences (``seqname''), although they have the same "
"gene IDs (``gene_id'') which must be globally unique",
gt_genome_node_get_line_number(first_node),
gt_genome_node_get_line_number(gn));
had_err = -1;
break;
}
}
if (!had_err) {
mRNA_node = gt_feature_node_new(mRNA_seqid, gt_ft_mRNA, mRNA_range.start,
mRNA_range.end, mRNA_strand);
if ((tname = gt_hashmap_get(cinfo->transcript_id_to_name_mapping,
(const char*) key))) {
gt_feature_node_add_attribute((GtFeatureNode*) mRNA_node, GT_GFF_NAME,
tname);
}
/* register children */
for (i = 0; i < gt_array_size(gt_genome_node_array); i++) {
gn = *(GtGenomeNode**) gt_array_get(gt_genome_node_array, i);
gt_feature_node_add_child((GtFeatureNode*) mRNA_node,
(GtFeatureNode*) gn);
}
/* store the mRNA */
gt_array_add(mRNAs, mRNA_node);
}
return had_err;
}
int gt_track_unit_test(GtError *err)
{
int had_err = 0;
GtBlock *b[4];
GtRange r[4];
GtTrack *track;
GtGenomeNode *parent[4], *gn[4];
GtStr *title;
double height, tmp;
GtStyle *sty;
unsigned long i;
GtLineBreaker *lb;
double t_rest = 0,
l_rest = 0;
gt_error_check(err);
title = gt_str_new_cstr("test");
r[0].start=100UL;
r[0].end=1000UL;
r[1].start=1001UL;
r[1].end=1500UL;
r[2].start=700UL;
r[2].end=1200UL;
r[3].start=10UL;
r[3].end=200UL;
for (i=0; i<4; i++)
{
parent[i] = gt_feature_node_new(title, gt_ft_gene, r[i].start, r[i].end,
GT_STRAND_FORWARD);
gn[i] = gt_feature_node_new(title, gt_ft_exon, r[i].start, r[i].end,
GT_STRAND_FORWARD);
gt_feature_node_add_child((GtFeatureNode*) parent[i],
(GtFeatureNode*) gn[i]);
gt_feature_node_add_attribute((GtFeatureNode*) parent[i], GT_GFF_NAME,
"parent");
gt_feature_node_add_attribute((GtFeatureNode*) gn[i], GT_GFF_NAME, "child");
}
for (i=0; i<4; i++)
{
b[i] = gt_block_new();
gt_block_set_range(b[i], r[i]);
gt_block_insert_element(b[i], (GtFeatureNode*) parent[i]);
gt_block_insert_element(b[i], (GtFeatureNode*) gn[i]);
}
lb = gt_line_breaker_bases_new();
sty = gt_style_new(err);
if (gt_style_get_num(sty, "format", "track_caption_font_size", &tmp,
NULL, err) == GT_STYLE_QUERY_NOT_SET) {
tmp = TEXT_SIZE_DEFAULT;
}
t_rest += tmp;
if (gt_style_get_num(sty, "format", "track_caption_space", &tmp,
NULL, err) == GT_STYLE_QUERY_NOT_SET) {
tmp = CAPTION_BAR_SPACE_DEFAULT;
}
t_rest += tmp;
if (gt_style_get_num(sty, "format", "track_vspace", &tmp,
NULL, err) == GT_STYLE_QUERY_NOT_SET) {
tmp = TRACK_VSPACE_DEFAULT;
}
t_rest += tmp;
if (gt_style_get_num(sty, "format", "bar_vspace", &l_rest,
NULL, err) == GT_STYLE_QUERY_NOT_SET) {
l_rest = BAR_VSPACE_DEFAULT;
}
track = gt_track_new(title, GT_UNDEF_ULONG, true, lb);
gt_ensure(had_err, track);
gt_ensure(had_err, gt_track_get_title(track) == title);
gt_ensure(had_err, gt_track_get_number_of_lines(track) == 0);
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest);
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_insert_block(track, b[0], err) == 0);
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_get_number_of_lines(track) == 1);
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest + l_rest + BAR_HEIGHT_DEFAULT);
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_insert_block(track, b[1], err) == 0);
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_get_number_of_lines(track) == 1);
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest + l_rest + BAR_HEIGHT_DEFAULT);
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_insert_block(track, b[2], err) == 0);
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_get_number_of_lines(track) == 2);
gt_ensure(had_err, gt_track_insert_block(track, b[3], err) == 0);
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_get_number_of_lines(track) == 2);
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest + 2*(l_rest + BAR_HEIGHT_DEFAULT));
gt_ensure(had_err, !gt_error_is_set(err));
gt_style_set_num(sty, "exon", "bar_height", 42);
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest + 2*(l_rest+42));
gt_ensure(had_err, !gt_error_is_set(err));
gt_style_set_num(sty, "gene", "bar_height", 23);
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest + 2*(l_rest+42));
gt_ensure(had_err, !gt_error_is_set(err));
gt_style_unset(sty, "exon", "bar_height");
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest + 2*(l_rest+23));
gt_ensure(had_err, !gt_error_is_set(err));
gt_style_unset(sty, "gene", "bar_height");
gt_style_set_num(sty, "format", "bar_height", 99);
gt_ensure(had_err, gt_track_get_height(track, &height, sty, err) == 0);
gt_ensure(had_err, height == t_rest + 2*(l_rest+99));
gt_ensure(had_err, !gt_error_is_set(err));
gt_ensure(had_err, gt_track_get_number_of_discarded_blocks(track) == 0);
gt_track_delete(track);
gt_str_delete(title);
gt_style_delete(sty);
for (i=0; i<4; i++)
{
gt_block_delete(b[i]);
gt_genome_node_delete(parent[i]);
}
return had_err;
}
static void infer_cds_visitor_infer_utrs(AgnInferCDSVisitor *v)
{
GtFeatureNode *start_codon, *stop_codon;
bool exonsexplicit = gt_array_size(v->exons) > 0;
bool cdsexplicit = gt_array_size(v->cds) > 0;
bool startcodon_check = gt_array_size(v->starts) == 1 &&
(start_codon = gt_array_get(v->starts, 0)) != NULL;
bool stopcodon_check = gt_array_size(v->stops) == 1 &&
(stop_codon = gt_array_get(v->stops, 0)) != NULL;
bool caninferutrs = exonsexplicit && startcodon_check && stopcodon_check;
if(gt_array_size(v->utrs) > 0)
{
return;
}
else if(!cdsexplicit && !caninferutrs)
{
return;
}
GtGenomeNode **leftcodon = gt_array_get(v->starts, 0);
GtGenomeNode **rightcodon = gt_array_get(v->stops, 0);
GtStrand strand = gt_feature_node_get_strand(v->mrna);
const char *lefttype = "five_prime_UTR";
const char *righttype = "three_prime_UTR";
if(strand == GT_STRAND_REVERSE)
{
lefttype = "three_prime_UTR";
righttype = "five_prime_UTR";
void *temp = leftcodon;
leftcodon = rightcodon;
rightcodon = temp;
}
GtRange leftrange = gt_genome_node_get_range(*leftcodon);
GtRange rightrange = gt_genome_node_get_range(*rightcodon);
GtUword i;
for(i = 0; i < gt_array_size(v->exons); i++)
{
GtGenomeNode **exon = gt_array_get(v->exons, i);
GtRange exonrange = gt_genome_node_get_range(*exon);
if(exonrange.start < leftrange.start)
{
GtRange utrrange;
if(gt_range_overlap(&exonrange, &leftrange))
{
utrrange.start = exonrange.start;
utrrange.end = leftrange.start - 1;
}
else
{
utrrange = exonrange;
}
GtGenomeNode *utr = gt_feature_node_new(gt_genome_node_get_seqid(*exon),
lefttype, utrrange.start,
utrrange.end, strand);
if(v->source)
gt_feature_node_set_source((GtFeatureNode *)utr, v->source);
gt_feature_node_add_child(v->mrna, (GtFeatureNode *)utr);
gt_array_add(v->utrs, utr);
}
if(exonrange.end > rightrange.end)
{
GtRange utrrange;
if(gt_range_overlap(&exonrange, &rightrange))
{
utrrange.start = rightrange.end + 1;
utrrange.end = exonrange.end;
}
else
{
utrrange = exonrange;
}
GtGenomeNode *utr = gt_feature_node_new(gt_genome_node_get_seqid(*exon),
righttype, utrrange.start,
utrrange.end, strand);
if(v->source)
gt_feature_node_set_source((GtFeatureNode *)utr, v->source);
gt_feature_node_add_child(v->mrna, (GtFeatureNode *)utr);
gt_array_add(v->utrs, utr);
}
}
}
static int construct_genes(GT_UNUSED void *key, void *value, void *data,
GtError *err)
{
GtHashmap *transcript_id_hash = (GtHashmap*) value;
ConstructionInfo *cinfo = (ConstructionInfo*) data;
GtQueue *genome_nodes = cinfo->genome_nodes;
const char *gname;
GtArray *mRNAs = gt_array_new(sizeof (GtGenomeNode*));
GtGenomeNode *gene_node, *gn;
GtStrand gene_strand;
GtRange gene_range;
GtStr *gene_seqid;
GtUword i;
int had_err = 0;
gt_error_check(err);
gt_assert(key && value && data);
cinfo->mRNAs = mRNAs;
had_err = gt_hashmap_foreach(transcript_id_hash, construct_mRNAs, cinfo, err);
if (!had_err) {
gt_assert(gt_array_size(mRNAs)); /* at least one mRNA constructed */
/* determine the range and the strand of the gene */
gn = *(GtGenomeNode**) gt_array_get(mRNAs, 0);
gene_range = gt_genome_node_get_range(gn);
gene_strand = gt_feature_node_get_strand((GtFeatureNode*) gn);
gene_seqid = gt_genome_node_get_seqid(gn);
for (i = 1; i < gt_array_size(mRNAs); i++) {
GtRange range;
gn = *(GtGenomeNode**) gt_array_get(mRNAs, i);
range = gt_genome_node_get_range(gn);
gene_range = gt_range_join(&gene_range, &range);
gene_strand = gt_strand_join(gene_strand,
gt_feature_node_get_strand((GtFeatureNode*) gn));
gt_assert(gt_str_cmp(gene_seqid, gt_genome_node_get_seqid(gn)) == 0);
}
gene_node = gt_feature_node_new(gene_seqid, gt_ft_gene, gene_range.start,
gene_range.end, gene_strand);
if ((gname = gt_hashmap_get(cinfo->gene_id_to_name_mapping,
(const char*) key))) {
gt_feature_node_add_attribute((GtFeatureNode*) gene_node, GT_GFF_NAME,
gname);
}
/* register children */
for (i = 0; i < gt_array_size(mRNAs); i++) {
gn = *(GtGenomeNode**) gt_array_get(mRNAs, i);
gt_feature_node_add_child((GtFeatureNode*) gene_node,
(GtFeatureNode*) gn);
}
/* store the gene */
gt_queue_add(genome_nodes, gene_node);
/* free */
gt_array_delete(mRNAs);
}
return had_err;
}