static void create_transitive_part_of_edges(GtTypeNode *node,
GtBoolMatrix *part_of_out_edges,
GtBoolMatrix *part_of_in_edges,
GtArray *node_stack)
{
unsigned long i, j;
if (gt_array_size(node_stack)) {
for (i = gt_bool_matrix_get_first_column(part_of_in_edges, node->num);
i != gt_bool_matrix_get_last_column(part_of_in_edges, node->num);
i = gt_bool_matrix_get_next_column(part_of_in_edges, node->num, i)) {
for (j = 0; j < gt_array_size(node_stack); j++) {
GtTypeNode *child = *(GtTypeNode**) gt_array_get(node_stack, j);
gt_bool_matrix_set(part_of_out_edges, i, child->num, true);
gt_bool_matrix_set(part_of_in_edges, child->num, i, true);
}
}
}
gt_array_add(node_stack, node);
for (i = 0; i < gt_array_size(node->is_a_out_edges); i++) {
GtTypeNode *parent = *(GtTypeNode**) gt_array_get(node->is_a_out_edges, i);
create_transitive_part_of_edges(parent, part_of_out_edges, part_of_in_edges,
node_stack);
}
gt_array_pop(node_stack);
}
static GtArray *gaeval_visitor_union(GtArray *cov1, GtArray *cov2)
{
agn_assert(cov1 && cov2);
gt_array_add_array(cov1, cov2);
if(gt_array_size(cov1) > 1)
gt_array_sort(cov1, (GtCompare)gt_range_compare);
GtArray *runion = gt_array_new(sizeof(GtRange));
if(gt_array_size(cov1) == 0)
return runion;
GtRange *rng = gt_array_get(cov1, 0);
gt_array_add(runion, *rng);
GtRange *prev = gt_array_get(runion, 0);
if(gt_array_size(cov1) == 1)
return runion;
GtUword i;
for(i = 1; i < gt_array_size(cov1); i++)
{
rng = gt_array_get(cov1, i);
if(gt_range_overlap(rng, prev))
*prev = gt_range_join(rng, prev);
else
{
gt_array_add(runion, *rng);
prev = gt_array_get(runion, gt_array_size(runion) - 1);
}
}
return runion;
}
int gt_feature_index_add_gff3file(GtFeatureIndex *feature_index,
const char *gff3file, GtError *err)
{
GtNodeStream *gff3_in_stream;
GtGenomeNode *gn;
GtArray *tmp;
int had_err = 0;
GtUword i;
gt_error_check(err);
gt_assert(feature_index && gff3file);
tmp = gt_array_new(sizeof (GtGenomeNode*));
gff3_in_stream = gt_gff3_in_stream_new_unsorted(1, &gff3file);
while (!(had_err = gt_node_stream_next(gff3_in_stream, &gn, err)) && gn)
gt_array_add(tmp, gn);
if (!had_err) {
GtNodeVisitor *feature_visitor = gt_feature_visitor_new(feature_index);
for (i=0;i<gt_array_size(tmp);i++) {
gn = *(GtGenomeNode**) gt_array_get(tmp, i);
/* no need to lock, add_*_node() is synchronized */
had_err = gt_genome_node_accept(gn, feature_visitor, NULL);
gt_assert(!had_err); /* cannot happen */
}
gt_node_visitor_delete(feature_visitor);
}
gt_node_stream_delete(gff3_in_stream);
for (i=0;i<gt_array_size(tmp);i++)
gt_genome_node_delete(*(GtGenomeNode**) gt_array_get(tmp, i));
gt_array_delete(tmp);
return had_err;
}
static void infer_cds_visitor_check_cds_multi(AgnInferCDSVisitor *v)
{
if(gt_array_size(v->cds) <= 1)
{
return;
}
GtFeatureNode **firstsegment = gt_array_get(v->cds, 0);
const char *id = gt_feature_node_get_attribute(*firstsegment, "ID");
if(id == NULL)
{
char newid[64];
sprintf(newid, "CDS%lu", v->cdscounter++);
gt_feature_node_add_attribute(*firstsegment, "ID", newid);
}
gt_feature_node_make_multi_representative(*firstsegment);
GtUword i;
for(i = 0; i < gt_array_size(v->cds); i++)
{
GtFeatureNode **segment = gt_array_get(v->cds, i);
if(!gt_feature_node_is_multi(*segment))
{
gt_feature_node_set_multi_representative(*segment, *firstsegment);
}
}
}
void gt_ranges_copy_to_opposite_strand(GtArray *outranges,
const GtArray *inranges,
GtUword gen_total_length,
GtUword gen_offset)
{
GtRange range;
GtUword i;
/* outranges are empty */
gt_assert(!gt_array_size(outranges));
/* inranges are not empty */
gt_assert(gt_array_size(inranges));
for (i = gt_array_size(inranges); i > 0; i--) {
/* genomic offset is defined */
gt_assert(gen_offset != GT_UNDEF_UWORD);
range.start = gen_total_length - 1
- (((GtRange*) gt_array_get(inranges, i-1))->end -
gen_offset)
+ gen_offset;
range.end = gen_total_length - 1
- (((GtRange*) gt_array_get(inranges, i-1))->start -
gen_offset)
+ gen_offset;
gt_array_add(outranges, range);
}
/* outranges has the same number of elements as inranges */
gt_assert(gt_array_size(inranges) == gt_array_size(outranges));
}
static void xml_outputSCRline(const GthAGS *ags, unsigned int indentlevel,
GtFile *outfp)
{
GthSpliceSiteProb *splicesiteprob;
unsigned long i;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<SCR_line>\n");
indentlevel++;
for (i = 0; i < gt_array_size(ags->exons) - 1; i++) {
splicesiteprob = (GthSpliceSiteProb*) gt_array_get(ags->splicesiteprobs, i);
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<exon-intron don_prob=\"%.3f\" "
"acc_prob=\"%.3f\" e_score=\"%.3f\"/>\n",
splicesiteprob->donorsiteprob,
splicesiteprob->acceptorsiteprob,
((GthExonAGS*) gt_array_get(ags->exons, i))->score);
}
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "<exon-only e_score=\"%.3f\"/>\n",
((GthExonAGS*) gt_array_get(ags->exons, i))->score);
indentlevel--;
gth_indent(outfp, indentlevel);
gt_file_xprintf(outfp, "</SCR_line>\n");
}
static void convert_chain_to_inverted_chain(GthInvertedChain *inverted_chain,
GthChain *chain)
{
unsigned long i, lastexonnum = gt_array_size(chain->forwardranges) - 1;
GtRange range;
/* inverted chain is empty */
gt_assert(!gt_array_size(inverted_chain->forwardranges));
/* chain is not empty */
gt_assert(gt_array_size(chain->forwardranges));
/* copy file and sequence numbers */
inverted_chain->gen_file_num = chain->gen_file_num;
inverted_chain->gen_seq_num = chain->gen_seq_num;
inverted_chain->ref_file_num = chain->ref_file_num;
inverted_chain->ref_seq_num = chain->ref_seq_num;
/* save startpos */
inverted_chain->startpos = ((GtRange*)
gt_array_get_first(chain->forwardranges))->start;
/* save endpos */
inverted_chain->endpos = ((GtRange*)
gt_array_get_last(chain->forwardranges))->end;
/* convert (potential) exons to (potential) introns */
for (i = 0; i < lastexonnum; i++) {
range.start = ((GtRange*) gt_array_get(chain->forwardranges, i))
->end + 1;
range.end = ((GtRange*) gt_array_get(chain->forwardranges, i+1))
->start - 1;
gt_array_add(inverted_chain->forwardranges, range);
}
}
static int gt_sort_stream_next(GtNodeStream *ns, GtGenomeNode **gn,
GtError *err)
{
GtSortStream *sort_stream;
GtGenomeNode *node, *eofn;
int had_err = 0;
gt_error_check(err);
sort_stream = gt_sort_stream_cast(ns);
if (!sort_stream->sorted) {
while (!(had_err = gt_node_stream_next(sort_stream->in_stream, &node,
err)) && node) {
if ((eofn = gt_eof_node_try_cast(node)))
gt_genome_node_delete(eofn); /* get rid of EOF nodes */
else
gt_array_add(sort_stream->nodes, node);
}
if (!had_err) {
gt_genome_nodes_sort_stable(sort_stream->nodes);
sort_stream->sorted = true;
}
}
if (!had_err) {
gt_assert(sort_stream->sorted);
if (sort_stream->idx < gt_array_size(sort_stream->nodes)) {
*gn = *(GtGenomeNode**) gt_array_get(sort_stream->nodes,
sort_stream->idx);
sort_stream->idx++;
/* join region nodes with the same sequence ID */
if (gt_region_node_try_cast(*gn)) {
GtRange range_a, range_b;
while (sort_stream->idx < gt_array_size(sort_stream->nodes)) {
node = *(GtGenomeNode**) gt_array_get(sort_stream->nodes,
sort_stream->idx);
if (!gt_region_node_try_cast(node) ||
gt_str_cmp(gt_genome_node_get_seqid(*gn),
gt_genome_node_get_seqid(node))) {
/* the next node is not a region node with the same ID */
break;
}
range_a = gt_genome_node_get_range(*gn);
range_b = gt_genome_node_get_range(node);
range_a = gt_range_join(&range_a, &range_b);
gt_genome_node_set_range(*gn, &range_a);
gt_genome_node_delete(node);
sort_stream->idx++;
}
}
return 0;
}
}
if (!had_err) {
gt_array_reset(sort_stream->nodes);
*gn = NULL;
}
return had_err;
}
static double gaeval_visitor_introns_confirmed(GtArray *introns, GtArray *gaps)
{
agn_assert(introns && gaps);
GtUword intron_count = gt_array_size(introns);
GtUword gap_count = gt_array_size(gaps);
agn_assert(intron_count > 0);
if(gap_count == 0)
return 0.0;
GtUword i, j, num_confirmed = 0;
for(i = 0; i < intron_count; i++)
{
GtGenomeNode *intron = *(GtGenomeNode **)gt_array_get(introns, i);
GtRange intron_range = gt_genome_node_get_range(intron);
for(j = 0; j < gap_count; j++)
{
GtGenomeNode *gap = *(GtGenomeNode **)gt_array_get(gaps, j);
GtRange gap_range = gt_genome_node_get_range(gap);
if(gt_range_compare(&intron_range, &gap_range) == 0)
{
num_confirmed++;
break;
}
}
}
return (double)num_confirmed / (double)intron_count;
}
static void snp_annotator_stream_free(GtNodeStream *ns)
{
GtUword i;
GtSNPAnnotatorStream *sas;
if (!ns) return;
sas = gt_snp_annotator_stream_cast(ns);
gt_region_mapping_delete(sas->rmap);
while (gt_queue_size(sas->snps) > 0) {
gt_genome_node_delete((GtGenomeNode*) gt_queue_get(sas->snps));
}
while (gt_queue_size(sas->outqueue) > 0) {
gt_genome_node_delete((GtGenomeNode*) gt_queue_get(sas->outqueue));
}
for (i = 0; i < gt_array_size(sas->instreams); i++) {
gt_node_stream_delete(*(GtNodeStream**) gt_array_get(sas->instreams, i));
}
for (i = 0; i < gt_array_size(sas->cur_gene_set); i++) {
gt_genome_node_delete(*(GtGenomeNode**) gt_array_get(sas->cur_gene_set, i));
}
gt_array_delete(sas->cur_gene_set);
gt_node_stream_delete(sas->merge_stream);
gt_array_delete(sas->instreams);
gt_queue_delete(sas->snps);
gt_queue_delete(sas->outqueue);
}
static void gv_test_calc_integrity(AgnUnitTest *test)
{
const char *filename = "data/gff3/gaeval-stream-unit-test-2.gff3";
GtNodeStream *align_in = gt_gff3_in_stream_new_unsorted(1, &filename);
AgnGaevalParams params = { 0.6, 0.3, 0.05, 0.05, 400, 200, 100 };
GtNodeVisitor *nv = agn_gaeval_visitor_new(align_in, params);
AgnGaevalVisitor *gv = gaeval_visitor_cast(nv);
gt_node_stream_delete(align_in);
GtNodeStream *gff3in = gt_gff3_in_stream_new_unsorted(1, &filename);
GtHashmap *typestokeep = gt_hashmap_new(GT_HASH_STRING, NULL, NULL);
gt_hashmap_add(typestokeep, "mRNA", "mRNA");
GtNodeStream *filtstream = agn_filter_stream_new(gff3in, typestokeep);
GtLogger *logger = gt_logger_new(true, "", stderr);
GtNodeStream *ics = agn_infer_cds_stream_new(filtstream, NULL, logger);
GtNodeStream *ies = agn_infer_exons_stream_new(ics, NULL, logger);
GtError *error = gt_error_new();
GtArray *feats = gt_array_new( sizeof(GtFeatureNode *) );
GtNodeStream *featstream = gt_array_out_stream_new(ies, feats, error);
int result = gt_node_stream_pull(featstream, error);
if(result == -1)
{
fprintf(stderr, "[AgnGaevalVisitor::gv_test_calc_integrity] error "
"processing GFF3: %s\n", gt_error_get(error));
return;
}
gt_node_stream_delete(gff3in);
gt_node_stream_delete(filtstream);
gt_node_stream_delete(featstream);
gt_node_stream_delete(ics);
gt_node_stream_delete(ies);
gt_logger_delete(logger);
gt_hashmap_delete(typestokeep);
agn_assert(gt_array_size(feats) == 2);
GtFeatureNode *g1 = *(GtFeatureNode **)gt_array_get(feats, 0);
GtFeatureNode *g2 = *(GtFeatureNode **)gt_array_get(feats, 1);
double cov1 = gaeval_visitor_calculate_coverage(gv, g1, error);
double cov2 = gaeval_visitor_calculate_coverage(gv, g2, error);
double int1 = gaeval_visitor_calculate_integrity(gv, g1, cov1, NULL, error);
double int2 = gaeval_visitor_calculate_integrity(gv, g2, cov2, NULL, error);
bool test1 = fabs(cov1 - 1.000) < 0.001 &&
fabs(cov2 - 0.997) < 0.001 &&
fabs(int1 - 0.850) < 0.001 &&
fabs(int2 - 0.863) < 0.001;
agn_unit_test_result(test, "calculate integrity", test1);
gt_error_delete(error);
gt_array_delete(feats);
gt_genome_node_delete((GtGenomeNode *)g1);
gt_genome_node_delete((GtGenomeNode *)g2);
gt_node_visitor_delete(nv);
}
static int gt_ltr_cluster_stream_next(GtNodeStream *ns,
GtGenomeNode **gn,
GtError *err)
{
GtLTRClusterStream *lcs;
GtGenomeNode *ref_gn;
int had_err = 0;
unsigned long i = 0;
gt_error_check(err);
lcs = gt_ltr_cluster_stream_cast(ns);
if (lcs->first_next) {
while (!(had_err = gt_node_stream_next(lcs->in_stream, gn, err)) && *gn) {
gt_assert(*gn && !had_err);
ref_gn = gt_genome_node_ref(*gn);
gt_array_add(lcs->nodes, ref_gn);
had_err = gt_genome_node_accept(*gn, (GtNodeVisitor*) lcs->lcv, err);
if (had_err) {
gt_genome_node_delete(*gn);
*gn = NULL;
break;
}
}
lcs->feat_to_encseq =
gt_ltr_cluster_prepare_seq_visitor_get_encseqs(lcs->lcv);
lcs->feat_to_encseq_keys =
gt_ltr_cluster_prepare_seq_visitor_get_features(lcs->lcv);
if (!had_err) {
for (i = 0; i < gt_str_array_size(lcs->feat_to_encseq_keys); i++) {
had_err = process_feature(lcs,
gt_str_array_get(lcs->feat_to_encseq_keys, i),
err);
if (had_err)
break;
}
}
if (!had_err) {
*gn = *(GtGenomeNode**) gt_array_get(lcs->nodes, lcs->next_index);
lcs->next_index++;
lcs->first_next = false;
return 0;
}
} else {
if (lcs->next_index >= gt_array_size(lcs->nodes))
*gn = NULL;
else {
*gn = *(GtGenomeNode**) gt_array_get(lcs->nodes, lcs->next_index);
lcs->next_index++;
}
return 0;
}
return had_err;
}
static void sort_matches_and_calc_buckets(GtArray *matches, GtArray *buckets,
GtUword *maxbucketlength)
{
GtUword i, currentstart = 0, currentend = 0;
GthMatch *matchptr;
Bucket bucket, *bucketptr;
gt_assert(gt_array_size(matches));
/* sort matches */
qsort(gt_array_get_space(matches), gt_array_size(matches), sizeof (GthMatch),
compare_matches);
/* init first bucket */
matchptr = gt_array_get_first(matches);
bucket.seqnum1 = matchptr->Storeseqnumreference;
bucket.seqnum2 = matchptr->Storeseqnumgenomic;
bucket.startpos = 0;
/* calc buckets */
for (i = 1; i < gt_array_size(matches); i++) {
matchptr = gt_array_get(matches, i);
if (matchptr->Storeseqnumreference != bucket.seqnum1 ||
matchptr->Storeseqnumgenomic != bucket.seqnum2) {
/* save the current bucket */
currentend = i - 1;
bucket.length = currentend - currentstart + 1;
gt_array_add(buckets, bucket);
/* create new bucket */
currentstart = i;
bucket.seqnum1 = matchptr->Storeseqnumreference;
bucket.seqnum2 = matchptr->Storeseqnumgenomic;
bucket.startpos = i;
}
}
/* save last bucket */
currentend = i - 1;
bucket.length = currentend - currentstart + 1;
gt_array_add(buckets, bucket);
/* compute maximum bucket length */
*maxbucketlength = 0;
for (i = 0; i < gt_array_size(buckets); i++) {
bucketptr = gt_array_get(buckets, i);
if (bucketptr->length > *maxbucketlength)
*maxbucketlength = bucketptr->length;
}
gt_assert(sum_of_bucket_lengths_equals_num_of_matches(buckets,
gt_array_size(matches)));
}
void agn_transcript_structure_gbk(GtFeatureNode *transcript, FILE *outstream)
{
gt_assert(transcript && outstream);
GtArray *exons = gt_array_new( sizeof(GtFeatureNode *) );
GtFeatureNodeIterator *iter = gt_feature_node_iterator_new_direct(transcript);
GtFeatureNode *child;
for
(
child = gt_feature_node_iterator_next(iter);
child != NULL;
child = gt_feature_node_iterator_next(iter)
)
{
if(agn_gt_feature_node_is_exon_feature(child))
gt_array_add(exons, child);
}
gt_feature_node_iterator_delete(iter);
gt_assert(gt_array_size(exons) > 0);
gt_array_sort(exons, (GtCompare)agn_gt_genome_node_compare);
if(gt_feature_node_get_strand(transcript) == GT_STRAND_REVERSE)
fputs("complement(", outstream);
if(gt_array_size(exons) == 1)
{
GtGenomeNode *exon = *(GtGenomeNode **)gt_array_get(exons, 0);
GtRange exonrange = gt_genome_node_get_range(exon);
fprintf(outstream, "<%lu..>%lu", exonrange.start, exonrange.end);
}
else
{
fputs("join(", outstream);
GtUword i;
for(i = 0; i < gt_array_size(exons); i++)
{
GtGenomeNode *exon = *(GtGenomeNode **)gt_array_get(exons, i);
GtRange exonrange = gt_genome_node_get_range(exon);
if(i == 0)
fprintf(outstream, "<%lu..%lu", exonrange.start, exonrange.end);
else if(i+1 == gt_array_size(exons))
fprintf(outstream, ",%lu..>%lu", exonrange.start, exonrange.end);
else
fprintf(outstream, ",%lu..%lu", exonrange.start, exonrange.end);
}
fputs(")", outstream);
}
if(gt_feature_node_get_strand(transcript) == GT_STRAND_REVERSE)
fputs(")", outstream);
}
static void gt_hmmer_model_hit_delete(GtHMMERModelHit *mh)
{
unsigned long i;
if (!mh) return;
for (i = 0; i < gt_array_size(mh->fwd_hits); i++)
gt_free(*(GtHMMERSingleHit**) gt_array_get(mh->fwd_hits, i));
gt_array_delete(mh->fwd_hits);
for (i = 0; i < gt_array_size(mh->rev_hits); i++)
gt_free(*(GtHMMERSingleHit**) gt_array_get(mh->rev_hits, i));
gt_array_delete(mh->rev_hits);
gt_free(mh);
}
static GtArray*
gaeval_visitor_intersect(GtGenomeNode *genemodel, GtGenomeNode *alignment)
{
agn_assert(genemodel && alignment);
GtFeatureNode *genefn = gt_feature_node_cast(genemodel);
GtFeatureNode *algnfn = gt_feature_node_cast(alignment);
agn_assert(gt_feature_node_has_type(genefn, "mRNA"));
GtStrand genestrand = gt_feature_node_get_strand(genefn);
GtStrand algnstrand = gt_feature_node_get_strand(algnfn);
if(genestrand != algnstrand)
return NULL;
GtArray *covered_parts = gt_array_new( sizeof(GtRange) );
GtArray *exons = agn_typecheck_select(genefn, agn_typecheck_exon);
GtWord i;
for(i = 0; i < gt_array_size(exons); i++)
{
GtGenomeNode *exon = *(GtGenomeNode **)gt_array_get(exons, i);
GtRange exonrange = gt_genome_node_get_range(exon);
GtFeatureNodeIterator *aniter = gt_feature_node_iterator_new(algnfn);
GtFeatureNode *tempaln;
GtRange nullrange = {0, 0};
for(tempaln = gt_feature_node_iterator_next(aniter);
tempaln != NULL;
tempaln = gt_feature_node_iterator_next(aniter))
{
if(gt_feature_node_has_type(tempaln, "match_gap"))
continue;
GtRange alnrange = gt_genome_node_get_range((GtGenomeNode *) tempaln);
GtRange intr = gaeval_visitor_range_intersect(&exonrange, &alnrange);
if(gt_range_compare(&intr, &nullrange) != 0)
gt_array_add(covered_parts, intr);
}
gt_feature_node_iterator_delete(aniter);
}
gt_array_delete(exons);
for(i = 0; i < gt_array_size(covered_parts); i++)
{
GtRange *r1 = gt_array_get(covered_parts, i);
GtUword j;
for(j = i+1; j < gt_array_size(covered_parts); j++)
{
GtRange *r2 = gt_array_get(covered_parts, j);
agn_assert(gt_range_overlap(r1, r2) == false);
}
}
return covered_parts;
}
bool gt_ranges_do_not_overlap(const GtArray *ranges)
{
GtUword i;
gt_assert(ranges && gt_array_size(ranges));
for (i = 1; i < gt_array_size(ranges); i++) {
if (gt_range_overlap(gt_array_get(ranges, i-1), gt_array_get(ranges, i)))
return false;
}
return true;
}
static int gt_ltrdigest_pdom_visitor_process_hit(GT_UNUSED void *key, void *val,
void *data,
GT_UNUSED GtError *err)
{
GtHMMERModelHit *mh = (GtHMMERModelHit*) val;
GtLTRdigestPdomVisitor *lv = (GtLTRdigestPdomVisitor*) data;
const char *mdl = (const char*) key;
GtArray *hits = NULL;
GtUword nof_hits;
GtFragment *frags;
if (gt_double_compare(mh->best_fwd, mh->best_rev) <= 0)
hits = mh->fwd_hits;
else
hits = mh->rev_hits;
gt_assert(hits);
nof_hits = gt_array_size(hits);
if (nof_hits == 0) return 0;
if (nof_hits > 1UL) {
GtUword i, chainno;
frags = gt_malloc((size_t) nof_hits * sizeof (GtFragment));
for (i = 0; i < nof_hits; i++) {
GtHMMERSingleHit *h = *(GtHMMERSingleHit**) gt_array_get(hits, i);
gt_assert(h);
frags[i].startpos1 = h->hmmfrom;
frags[i].endpos1 = h->hmmto;
frags[i].startpos2 = h->alifrom;
frags[i].endpos2 = h->alito;
frags[i].weight = (GtWord) (h->alito - h->alifrom + 1) * h->score;
frags[i].data = h;
}
qsort(frags, (size_t) nof_hits, sizeof (GtFragment),
gt_ltrdigest_pdom_visitor_fragcmp);
gt_log_log("%s: chaining "GT_WU" frags", mdl, nof_hits);
gt_globalchaining_max(frags, nof_hits,
(GtUword) lv->chain_max_gap_length,
gt_ltrdigest_pdom_visitor_chainproc, &chainno);
gt_free(frags);
for (i = 0; i < nof_hits; i++) {
GtHMMERSingleHit *h = *(GtHMMERSingleHit**) gt_array_get(hits, i);
(void) gt_ltrdigest_pdom_visitor_attach_hit(lv, mh, h);
}
} else {
GtUword chainno = 0UL;
GtHMMERSingleHit *h = *(GtHMMERSingleHit**) gt_array_get(hits, 0);
gt_array_add(h->chains, chainno);
(void) gt_ltrdigest_pdom_visitor_attach_hit(lv, mh, h);
}
return 0;
}
static int get_next_free_line(GtTrack *track, GtLine **result, GtBlock *block,
GtError *err)
{
unsigned long i;
GtLine* line;
int had_err = 0;
bool is_occupied;
gt_assert(track);
/* find unoccupied line -- may need optimisation */
for (i = 0; i < gt_array_size(track->lines); i++) {
line = *(GtLine**) gt_array_get(track->lines, i);
had_err = gt_line_breaker_line_is_occupied(track->lb, &is_occupied, line,
block, err);
if (had_err)
break;
if (!is_occupied) {
*result = line;
return 0;
}
}
/* all lines are occupied, we need o create a new one */
if (!had_err) {
/* if line limit is hit, do not create any more lines! */
if (track->max_num_lines != GT_UNDEF_ULONG
&& gt_array_size(track->lines) == track->max_num_lines)
{
track->discarded_blocks++;
*result = NULL;
}
/* make sure there is only one line if 'split_lines' is set to false */
if (!track->split)
{
if (gt_array_size(track->lines) < 1)
{
line = gt_line_new();
gt_array_add(track->lines, line);
}
else
line = *(GtLine**) gt_array_get(track->lines, 0);
gt_assert(gt_array_size(track->lines) == 1);
}
else
{
line = gt_line_new();
gt_array_add(track->lines, line);
}
gt_assert(line);
}
*result = line;
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 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);
}
}
}
bool gt_ranges_are_sorted(const GtArray *ranges)
{
GtUword i;
gt_assert(ranges);
for (i = 1; i < gt_array_size(ranges); i++) {
if (gt_range_compare(gt_array_get(ranges, i-1),
gt_array_get(ranges, i)) == 1) {
return false;
}
}
return true;
}
static int gtf_show_transcript(GtFeatureNode *feature_node,
GtGTFVisitor *gtf_visitor, GtError *err)
{
GtFeatureNode *fn;
GtUword i;
int had_err;
gt_error_check(err);
gt_assert(feature_node && gtf_visitor);
gt_array_reset(gtf_visitor->exon_features);
gt_array_reset(gtf_visitor->CDS_features);
had_err = gt_feature_node_traverse_direct_children(feature_node, gtf_visitor,
save_exon_node, err);
if (gt_array_size(gtf_visitor->exon_features)) {
/* sort exon features */
qsort(gt_array_get_space(gtf_visitor->exon_features),
gt_array_size(gtf_visitor->exon_features), sizeof (GtGenomeNode*),
(GtCompare) gt_genome_node_compare);
/* show exon features */
gtf_visitor->transcript_id++;
for (i = 0; i < gt_array_size(gtf_visitor->exon_features); i++) {
fn = *(GtFeatureNode**) gt_array_get(gtf_visitor->exon_features, i);
gt_gff3_output_leading(fn, gtf_visitor->outfp);
gt_file_xprintf(gtf_visitor->outfp, "gene_id \""GT_WU"\"; transcript_id "
"\""GT_WU"."GT_WU"\";\n", gtf_visitor->gene_id,
gtf_visitor->gene_id, gtf_visitor->transcript_id);
}
}
if (gt_array_size(gtf_visitor->CDS_features)) {
/* sort CDS features */
qsort(gt_array_get_space(gtf_visitor->CDS_features),
gt_array_size(gtf_visitor->CDS_features), sizeof (GtGenomeNode*),
(GtCompare) gt_genome_node_compare);
/* show start_codon feature */
/* fn = *(GtFeatureNode**) */ (void) gt_array_get(gtf_visitor->CDS_features,
0);
/* XXX: to be done */
/* show CDS features */
for (i = 0; i < gt_array_size(gtf_visitor->CDS_features); i++) {
fn = *(GtFeatureNode**) gt_array_get(gtf_visitor->CDS_features, i);
gt_gff3_output_leading(fn, gtf_visitor->outfp);
gt_file_xprintf(gtf_visitor->outfp, "gene_id \""GT_WU"\"; transcript_id "
"\""GT_WU"."GT_WU"\";\n", gtf_visitor->gene_id,
gtf_visitor->gene_id, gtf_visitor->transcript_id);
}
/* XXX: show stop_codon feature and shorten last CDS feature */
}
return had_err;
}
static int gt_cluster_matches(GtClusteredSet *cs,
GtMatchEdgeTable *matchedgetab,
GtError *err)
{
unsigned long i;
for (i = 0; i < matchedgetab->num_of_edges; i++) {
if (gt_clustered_set_merge_clusters(cs,
((GtMatchEdge*)(gt_array_get(matchedgetab->edges, i)))->matchnum0,
((GtMatchEdge*)(gt_array_get(matchedgetab->edges, i)))->matchnum1,
err) != 0) {
return -1;
}
}
return 0;
}
void agn_bron_kerbosch( GtArray *R, GtArray *P, GtArray *X, GtArray *cliques,
bool skipsimplecliques )
{
gt_assert(R != NULL && P != NULL && X != NULL && cliques != NULL);
if(gt_array_size(P) == 0 && gt_array_size(X) == 0)
{
if(skipsimplecliques == false || gt_array_size(R) != 1)
{
GtUword i;
AgnTranscriptClique *clique = agn_transcript_clique_new();
for(i = 0; i < gt_array_size(R); i++)
{
GtFeatureNode *transcript = *(GtFeatureNode **)gt_array_get(R, i);
agn_transcript_clique_add(clique, transcript);
}
gt_array_add(cliques, clique);
}
}
while(gt_array_size(P) > 0)
{
GtGenomeNode *v = *(GtGenomeNode **)gt_array_get(P, 0);
// newR = R \union {v}
GtArray *newR = agn_gt_array_copy(R, sizeof(GtGenomeNode *));
gt_array_add(newR, v);
// newP = P \intersect N(v)
GtArray *newP = agn_feature_neighbors(v, P);
// newX = X \intersect N(v)
GtArray *newX = agn_feature_neighbors(v, X);
// Recursive call
// agn_bron_kerbosch(R \union {v}, P \intersect N(v), X \intersect N(X))
agn_bron_kerbosch(newR, newP, newX, cliques, skipsimplecliques);
// Delete temporary arrays just created
gt_array_delete(newR);
gt_array_delete(newP);
gt_array_delete(newX);
// P := P \ {v}
gt_array_rem(P, 0);
// X := X \union {v}
gt_array_add(X, v);
}
}
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 void potentialintronspostpro(GtArray *intronstoprocess,
unsigned long icdelta,
unsigned long icminremintronlength)
{
GtArray *originalintrons;
GtRange potintron;
unsigned long i, potintronlength,
minintronlength = 2 * icdelta + icminremintronlength;
originalintrons = gt_array_new(sizeof (GtRange));
/* save all (potential) introns */
gt_array_add_array(originalintrons, intronstoprocess);
/* reset introns to process */
gt_array_set_size(intronstoprocess, 0);
/* store introns */
for (i = 0; i < gt_array_size(originalintrons); i++) {
potintron = *(GtRange*) gt_array_get(originalintrons, i);
potintronlength = potintron.end - potintron.start + 1;
if (potintronlength >= minintronlength) {
/* keep this intron (plus/minus intron deltas)
that is, this intron is cut out later */
potintron.start += icdelta;
potintron.end -= icdelta;
gt_array_add(intronstoprocess, potintron);
}
/* else: skip this intron
that is, this intron is not cut out later */
}
gt_array_delete(originalintrons);
}
static int seqid_info_get(SeqidInfo *seqid_info, GtUword *seqnum,
GtUword *filenum, GtRange *outrange,
const GtRange *inrange,
GT_UNUSED const char *filename,
const char *seqid, GtError *err)
{
SeqidInfoElem *seqid_info_elem;
GtUword i;
gt_error_check(err);
gt_assert(seqid_info && seqnum && outrange && inrange);
for (i = 0; i < gt_array_size(seqid_info); i++) {
seqid_info_elem = gt_array_get(seqid_info, i);
if (seqid_info_elem->descrange.end == GT_UNDEF_UWORD ||
gt_range_contains(&seqid_info_elem->descrange, inrange)) {
*seqnum = seqid_info_elem->seqnum;
*filenum = seqid_info_elem->filenum;
*outrange = seqid_info_elem->descrange;
return 0;
}
}
gt_error_set(err,
"cannot find a sequence with ID \"%s\" "
"{range " GT_WU "," GT_WU ")",
seqid, inrange->start, inrange->end);
return -1;
}
bool gt_ranges_are_consecutive(const GtArray *ranges)
{
GtUword i;
for (i = 0; i < gt_array_size(ranges); i++) {
gt_assert(((GtRange*) gt_array_get(ranges, i))->start <=
((GtRange*) gt_array_get(ranges, i))->end);
if (i) {
/* check if ranges are consecutive */
if (((GtRange*) gt_array_get(ranges, i-1))->end >=
((GtRange*) gt_array_get(ranges, i))->start) {
return false;
}
}
}
return true;
}
static GtArray* generic_ranges_uniq(GtArray *out_ranges,
const GtArray *in_ranges, bool count)
{
GtUword i, *ctr_ptr, ctr = 1;
GtArray *count_array = NULL;
GtRange cur = { GT_UNDEF_UWORD, GT_UNDEF_UWORD },
prev = { GT_UNDEF_UWORD, GT_UNDEF_UWORD };
gt_assert(out_ranges && in_ranges);
gt_assert(gt_ranges_are_sorted(in_ranges));
if (count)
count_array = gt_array_new(sizeof (GtUword));
for (i = 0; i < gt_array_size(in_ranges); i++) {
cur = *(GtRange*) gt_array_get(in_ranges, i);
if (!i) {
gt_array_add(out_ranges, cur);
if (count)
gt_array_add(count_array, ctr);
}
else {
if (prev.start == cur.start && prev.end == cur.end) {
if (count) {
ctr_ptr = gt_array_get_last(count_array);
(*ctr_ptr)++;
}
}
else {
gt_array_add(out_ranges, cur);
if (count)
gt_array_add(count_array, ctr);
}
}
prev = cur;
}
return count_array;
}
