static int check_boundaries_visitor_check_rec(GtFeatureNode *parent,
GtFeatureNode *child,
GtError *err)
{
GtFeatureNodeIterator *fni;
GtFeatureNode *node;
GtRange range,
p_range;
int had_err = 0;
range = gt_genome_node_get_range((GtGenomeNode*) child);
p_range = gt_genome_node_get_range((GtGenomeNode*) parent);
if (range.start < p_range.start || range.end > p_range.end) {
gt_warning("%s child range " GT_WU "-" GT_WU " (file %s, line %u) not "
"contained in %s parent range " GT_WU "-" GT_WU " (file %s, "
"line %u)",
gt_feature_node_get_type(child),
range.start, range.end,
gt_genome_node_get_filename((GtGenomeNode*) child),
gt_genome_node_get_line_number((GtGenomeNode*) child),
gt_feature_node_get_type(parent),
p_range.start, p_range.end,
gt_genome_node_get_filename((GtGenomeNode*) parent),
gt_genome_node_get_line_number((GtGenomeNode*) parent));
}
fni = gt_feature_node_iterator_new_direct(child);
while ((node = gt_feature_node_iterator_next(fni))) {
had_err = check_boundaries_visitor_check_rec(child, node, err);
}
gt_feature_node_iterator_delete(fni);
return had_err;
}
static int gt_ltr_input_check_visitor_feature_node(GtNodeVisitor *nv,
GtFeatureNode *fn,
GtError *err)
{
GT_UNUSED GtLTRInputCheckVisitor *lv;
GtFeatureNodeIterator *fni;
bool seen_left = false;
GtFeatureNode *curnode = NULL,
*ltr_retrotrans = NULL,
*lltr = NULL,
*rltr = NULL;
int had_err = 0;
lv = gt_ltr_input_check_visitor_cast(nv);
gt_assert(lv);
gt_error_check(err);
/* traverse annotation subgraph and find LTR components */
fni = gt_feature_node_iterator_new(fn);
while (!had_err && (curnode = gt_feature_node_iterator_next(fni))) {
if (strcmp(gt_feature_node_get_type(curnode),
gt_ft_LTR_retrotransposon) == 0) {
ltr_retrotrans = curnode;
}
if (strcmp(gt_feature_node_get_type(curnode),
gt_ft_long_terminal_repeat) == 0) {
if (seen_left)
rltr = curnode;
else {
lltr = curnode;
seen_left = true;
}
}
}
gt_feature_node_iterator_delete(fni);
if (lv->only_ltrs) {
if (!had_err && !ltr_retrotrans) {
gt_error_set(err, "connected component with %s entry node (%s, line %u) "
"does not contain a '%s' node, which is required",
gt_feature_node_get_type(fn),
gt_genome_node_get_filename((GtGenomeNode*) fn),
gt_genome_node_get_line_number((GtGenomeNode*) fn),
gt_ft_LTR_retrotransposon);
had_err = -1;
}
}
if (!had_err && ltr_retrotrans && (!lltr || !rltr)) {
gt_error_set(err, "LTR_retrotransposon feature (%s, line %u) "
"does not contain two %s child features, both of which "
"are required",
gt_genome_node_get_filename((GtGenomeNode*) ltr_retrotrans),
gt_genome_node_get_line_number((GtGenomeNode*) ltr_retrotrans),
gt_ft_long_terminal_repeat);
had_err = -1;
}
return had_err;
}
static void default_track_selector(GtBlock *block, GtStr *result,
GT_UNUSED void *data)
{
GtGenomeNode *top;
char *basename;
gt_assert(block && result);
gt_str_reset(result);
top = (GtGenomeNode*) gt_block_get_top_level_feature(block);
/* we take the basename of the filename to have nicer output in the
generated graphic. this might lead to ``collapsed'' tracks, if two files
with different paths have the same basename. */
basename = gt_basename(gt_genome_node_get_filename(top));
gt_str_append_cstr(result, basename);
gt_free(basename);
gt_str_append_char(result, GT_FILENAME_TYPE_SEPARATOR);
gt_str_append_cstr(result, gt_block_get_type(block));
}
static int gtf_show_feature_node(GtFeatureNode *fn, void *data, GtError *err)
{
GtGTFVisitor *gtf_visitor = (GtGTFVisitor*) data;
int had_err = 0;
if (gt_feature_node_has_type(fn, gt_ft_gene)) {
gtf_visitor->gene_id++;
gtf_visitor->transcript_id = 0;
had_err = gtf_show_transcript(fn, gtf_visitor, err);
}
else if (gt_feature_node_has_type(fn, gt_ft_mRNA)) {
had_err = gtf_show_transcript(fn, gtf_visitor, err);
}
else if (!(gt_feature_node_has_type(fn, gt_ft_CDS) ||
gt_feature_node_has_type(fn, gt_ft_exon))) {
gt_warning("skipping GFF3 feature of type \"%s\" (from line %u in file "
"\"%s\")",
gt_feature_node_get_type(fn),
gt_genome_node_get_line_number((GtGenomeNode*) fn),
gt_genome_node_get_filename((GtGenomeNode*) fn));
}
return had_err;
}
static int buffer_is_sorted(void **elem, void *info, GtError *err)
{
GtGenomeNode *current_node, **last_node;
gt_error_check(err);
gt_assert(elem && info);
current_node = *(GtGenomeNode**) elem,
last_node = info;
if (*last_node && gt_genome_node_compare(last_node, ¤t_node) > 0) {
gt_assert(*last_node);
gt_error_set(err, "the file %s is not sorted (example: line %u and %u)",
gt_genome_node_get_filename(*last_node),
gt_genome_node_get_line_number(*last_node),
gt_genome_node_get_line_number(current_node));
return -1;
}
else
*last_node = current_node;
return 0;
}
static int add_ids_visitor_region_node(GtNodeVisitor *nv, GtRegionNode *rn,
GT_UNUSED GtError *err)
{
GtAddIDsVisitor *aiv;
const char *seqid;
int had_err = 0;
gt_error_check(err);
aiv = add_ids_visitor_cast(nv);
seqid = gt_str_get(gt_genome_node_get_seqid((GtGenomeNode*) rn));
if (gt_hashmap_get(aiv->undefined_sequence_regions, seqid)) {
gt_error_set(err, "genome feature with id \"%s\" has been defined before "
"the corresponding \"%s\" definition on line %u in file "
"\"%s\"", seqid, GT_GFF_SEQUENCE_REGION,
gt_genome_node_get_line_number((GtGenomeNode*) rn),
gt_genome_node_get_filename((GtGenomeNode*) rn));
had_err = -1;
}
if (!had_err) {
if (!gt_cstr_table_get(aiv->defined_seqids, seqid))
gt_cstr_table_add(aiv->defined_seqids, seqid);
gt_queue_add(aiv->node_buffer, rn);
}
return had_err;
}
static int check_cds_phases(GtArray *cds_features, GtCDSCheckVisitor *v,
bool is_multi, bool second_pass, GtError *err)
{
GtPhase current_phase, correct_phase = GT_PHASE_ZERO;
GtFeatureNode *fn;
GtStrand strand;
unsigned long i, current_length;
int had_err = 0;
gt_error_check(err);
gt_assert(cds_features);
gt_assert(gt_array_size(cds_features));
fn = *(GtFeatureNode**) gt_array_get_first(cds_features);
strand = gt_feature_node_get_strand(fn);
if (strand == GT_STRAND_REVERSE)
gt_array_reverse(cds_features);
for (i = 0; !had_err && i < gt_array_size(cds_features); i++) {
fn = *(GtFeatureNode**) gt_array_get(cds_features, i);
/* the first phase can be anything (except being undefined), because the
GFF3 spec says:
NOTE 4 - CDS features MUST have have a defined phase field. Otherwise it
is not possible to infer the correct polypeptides corresponding to
partially annotated genes. */
if ((!i && gt_feature_node_get_phase(fn) == GT_PHASE_UNDEFINED) ||
(i && gt_feature_node_get_phase(fn) != correct_phase)) {
if (gt_hashmap_get(v->cds_features, fn)) {
if (v->tidy && !is_multi && !gt_feature_node_has_children(fn)) {
/* we can split the feature */
gt_warning("%s feature on line %u in file \"%s\" has multiple "
"parents which require different phases; split feature",
gt_ft_CDS,
gt_genome_node_get_line_number((GtGenomeNode*) fn),
gt_genome_node_get_filename((GtGenomeNode*) fn));
gt_hashmap_add(v->cds_features_to_split, fn, fn);
v->splitting_is_necessary = true; /* split later */
}
else {
gt_error_set(err, "%s feature on line %u in file \"%s\" has multiple "
"parents which require different phases",
gt_ft_CDS,
gt_genome_node_get_line_number((GtGenomeNode*) fn),
gt_genome_node_get_filename((GtGenomeNode*) fn));
had_err = -1;
}
}
else {
if (v->tidy) {
if (!second_pass) {
gt_warning("%s feature on line %u in file \"%s\" has the wrong "
"phase %c -> correcting it to %c", gt_ft_CDS,
gt_genome_node_get_line_number((GtGenomeNode*) fn),
gt_genome_node_get_filename((GtGenomeNode*) fn),
GT_PHASE_CHARS[gt_feature_node_get_phase(fn)],
GT_PHASE_CHARS[correct_phase]);
}
gt_feature_node_set_phase(fn, correct_phase);
}
else {
gt_error_set(err, "%s feature on line %u in file \"%s\" has the "
"wrong phase %c (should be %c)", gt_ft_CDS,
gt_genome_node_get_line_number((GtGenomeNode*) fn),
gt_genome_node_get_filename((GtGenomeNode*) fn),
GT_PHASE_CHARS[gt_feature_node_get_phase(fn)],
GT_PHASE_CHARS[correct_phase]);
had_err = -1;
}
}
}
if (!had_err) {
current_phase = gt_feature_node_get_phase(fn);
current_length = gt_genome_node_get_length((GtGenomeNode*) fn);
correct_phase = (3 - (current_length - current_phase) % 3) % 3;
gt_hashmap_add(v->cds_features, fn, fn); /* record CDS feature */
}
}
return had_err;
}
static int genome_node_lua_get_filename(lua_State *L)
{
GtGenomeNode **gn = check_genome_node(L, 1);
lua_pushstring(L, gt_genome_node_get_filename(*gn));
return 1;
}
static int add_ids_visitor_feature_node(GtNodeVisitor *nv, GtFeatureNode *fn,
GtError *err)
{
AutomaticSequenceRegion *auto_sr;
GtAddIDsVisitor *aiv;
const char *seqid;
bool is_circular;
aiv = add_ids_visitor_cast(nv);
seqid = gt_str_get(gt_genome_node_get_seqid((GtGenomeNode*) fn));
if (aiv->ensure_sorting && !gt_cstr_table_get(aiv->defined_seqids, seqid)) {
gt_error_set(err, "the file %s is not sorted (seqid \"%s\" on line %u has "
"not been previously introduced with a \"%s\" line)",
gt_genome_node_get_filename((GtGenomeNode*) fn), seqid,
gt_genome_node_get_line_number((GtGenomeNode*) fn),
GT_GFF_SEQUENCE_REGION);
return -1;
}
if (!gt_cstr_table_get(aiv->defined_seqids, seqid)) {
GtFeatureNodeIterator *fni;
GtFeatureNode *node;
GtRange range = gt_genome_node_get_range((GtGenomeNode*) fn);
is_circular = gt_feature_node_get_attribute(fn, GT_GFF_IS_CIRCULAR)
? true : false;
if (!is_circular) {
fni = gt_feature_node_iterator_new(fn);
while ((node = gt_feature_node_iterator_next(fni))) {
GtRange node_range = gt_genome_node_get_range((GtGenomeNode*) node);
range = gt_range_join(&range, &node_range);
}
gt_feature_node_iterator_delete(fni);
}
/* sequence region has not been previously introduced -> check if one has
already been created automatically */
auto_sr = gt_hashmap_get(aiv->undefined_sequence_regions, seqid);
if (!auto_sr) {
GtStr *seqid_str;
/* sequence region has not been createad automatically -> do it now */
gt_warning("seqid \"%s\" on line %u in file \"%s\" has not been "
"previously introduced with a \"%s\" line, create such a line "
"automatically", seqid,
gt_genome_node_get_line_number((GtGenomeNode*) fn),
gt_genome_node_get_filename((GtGenomeNode*) fn),
GT_GFF_SEQUENCE_REGION);
auto_sr = automatic_sequence_region_new(is_circular);
seqid_str = gt_genome_node_get_seqid((GtGenomeNode*) fn);
auto_sr->sequence_region = gt_region_node_new(seqid_str, range.start,
range.end);
gt_hashmap_add(aiv->undefined_sequence_regions, gt_str_get(seqid_str),
auto_sr);
}
else {
if (auto_sr->is_circular) {
gt_assert(!is_circular); /* XXX */
}
else if (is_circular) {
gt_assert(!auto_sr->is_circular); /* XXX */
auto_sr->is_circular = true;
gt_genome_node_set_range(auto_sr->sequence_region, &range);
}
else {
GtRange joined_range,
sr_range = gt_genome_node_get_range(auto_sr->sequence_region);
/* update the range of the sequence region */
joined_range = gt_range_join(&range, &sr_range);
gt_genome_node_set_range(auto_sr->sequence_region, &joined_range);
}
}
gt_array_add(auto_sr->feature_nodes, fn);
}
else
gt_queue_add(aiv->node_buffer, fn);
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;
}
