static int gff3_visitor_region_node(GtNodeVisitor *nv, GtRegionNode *rn,
GT_UNUSED GtError *err)
{
GtGFF3Visitor *gff3_visitor;
gt_error_check(err);
gff3_visitor = gff3_visitor_cast(nv);
gt_assert(nv && rn);
gff3_version_string(nv);
if (!gff3_visitor->outstr) {
gt_file_xprintf(gff3_visitor->outfp, "%s %s "GT_WU" "GT_WU"\n",
GT_GFF_SEQUENCE_REGION,
gt_str_get(gt_genome_node_get_seqid((GtGenomeNode*) rn)),
gt_genome_node_get_start((GtGenomeNode*) rn),
gt_genome_node_get_end((GtGenomeNode*) rn));
} else {
gt_str_append_cstr(gff3_visitor->outstr, GT_GFF_SEQUENCE_REGION);
gt_str_append_cstr(gff3_visitor->outstr, " ");
gt_str_append_cstr(gff3_visitor->outstr,
gt_str_get(gt_genome_node_get_seqid((GtGenomeNode*) rn)));
gt_str_append_char(gff3_visitor->outstr, ' ');
gt_str_append_ulong(gff3_visitor->outstr,
gt_genome_node_get_start((GtGenomeNode*) rn));
gt_str_append_char(gff3_visitor->outstr, ' ');
gt_str_append_ulong(gff3_visitor->outstr,
gt_genome_node_get_end((GtGenomeNode*) rn));
gt_str_append_char(gff3_visitor->outstr, '\n');
}
return 0;
}
void gt_gff3_output_leading_str(GtFeatureNode *fn, GtStr *outstr)
{
GtGenomeNode *gn;
gt_assert(fn && outstr);
gn = (GtGenomeNode*) fn;
gt_str_append_str(outstr, gt_genome_node_get_seqid(gn));
gt_str_append_char(outstr, '\t');
gt_str_append_cstr(outstr, gt_feature_node_get_source(fn));
gt_str_append_char(outstr, '\t');
gt_str_append_cstr(outstr, gt_feature_node_get_type(fn));
gt_str_append_char(outstr, '\t');
gt_str_append_uword(outstr, gt_genome_node_get_start(gn));
gt_str_append_char(outstr, '\t');
gt_str_append_uword(outstr, gt_genome_node_get_end(gn));
gt_str_append_char(outstr, '\t');
if (gt_feature_node_score_is_defined(fn)) {
char buf[BUFSIZ];
(void) snprintf(buf, BUFSIZ, "%.3g", gt_feature_node_get_score(fn));
gt_str_append_cstr(outstr, buf);
} else
gt_str_append_char(outstr, '.');
gt_str_append_char(outstr, '\t');
gt_str_append_char(outstr, GT_STRAND_CHARS[gt_feature_node_get_strand(fn)]);
gt_str_append_char(outstr, '\t');
gt_str_append_char(outstr, GT_PHASE_CHARS[gt_feature_node_get_phase(fn)]);
gt_str_append_char(outstr, '\t');
}
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_set_utrs(AgnInferCDSVisitor *v)
{
GtGenomeNode **start;
GtUword i, cds_start;
if(!v->starts || gt_array_size(v->starts) != 1)
return;
start = gt_array_get(v->starts, 0);
cds_start = gt_genome_node_get_start(*start);
for(i = 0; i < gt_array_size(v->utrs); i++)
{
GtFeatureNode *utr = *(GtFeatureNode **)gt_array_get(v->utrs, i);
GtStrand strand = gt_feature_node_get_strand(utr);
GtUword utr_start = gt_genome_node_get_start((GtGenomeNode *)utr);
if(!gt_feature_node_has_type(utr, "five_prime_UTR") &&
!gt_feature_node_has_type(utr, "three_prime_UTR"))
{
if(strand == GT_STRAND_FORWARD)
{
if(utr_start < cds_start)
gt_feature_node_set_type(utr, "five_prime_UTR");
else
gt_feature_node_set_type(utr, "three_prime_UTR");
}
else
{
if(utr_start < cds_start)
gt_feature_node_set_type(utr, "three_prime_UTR");
else
gt_feature_node_set_type(utr, "five_prime_UTR");
}
}
}
}
static int gff3_visitor_region_node(GtNodeVisitor *nv, GtRegionNode *rn,
GT_UNUSED GtError *err)
{
GtGFF3Visitor *gff3_visitor;
gt_error_check(err);
gff3_visitor = gff3_visitor_cast(nv);
gt_assert(nv && rn);
gff3_version_string(nv);
gt_file_xprintf(gff3_visitor->outfp, "%s %s %lu %lu\n",
GT_GFF_SEQUENCE_REGION,
gt_str_get(gt_genome_node_get_seqid((GtGenomeNode*) rn)),
gt_genome_node_get_start((GtGenomeNode*) rn),
gt_genome_node_get_end((GtGenomeNode*) rn));
return 0;
}
void gt_gff3_output_leading(GtFeatureNode *fn, GtFile *outfp)
{
GtGenomeNode *gn;
gt_assert(fn);
gn = (GtGenomeNode*) fn;
gt_file_xprintf(outfp, "%s\t%s\t%s\t"GT_WU"\t"GT_WU"\t",
gt_str_get(gt_genome_node_get_seqid(gn)),
gt_feature_node_get_source(fn),
gt_feature_node_get_type(fn),
gt_genome_node_get_start(gn),
gt_genome_node_get_end(gn));
if (gt_feature_node_score_is_defined(fn))
gt_file_xprintf(outfp, "%.3g", gt_feature_node_get_score(fn));
else
gt_file_xfputc('.', outfp);
gt_file_xprintf(outfp, "\t%c\t%c\t",
GT_STRAND_CHARS[gt_feature_node_get_strand(fn)],
GT_PHASE_CHARS[gt_feature_node_get_phase(fn)]);
}
static int CpGIOverlap_stream_next(GtNodeStream * ns,
GtGenomeNode ** gn,
GtError * err)
{
GtGenomeNode * cur_node, * next_node;
GtFeatureNodeIterator * iter;
int err_num = 0;
*gn = NULL;
CpGIOverlap_stream * context;
const char * gene_name = NULL;
const char * overlap_name = NULL;
char chr_str[255];
int chr_num;
unsigned int TSS;
float CpGIOverlap;
context = CpGIOverlap_stream_cast(ns);
// find the genes, determine expression level
if(!gt_node_stream_next(context->in_stream,
&cur_node,
err
) && cur_node != NULL
)
{
*gn = cur_node;
// try casting as a feature node so we can test type
if(!gt_genome_node_try_cast(gt_feature_node_class(), cur_node))
{
return 0;
}
else // we found a feature node
{
// first check if it is a pseudo node, if so find the gene in it if available
if (gt_feature_node_is_pseudo(cur_node))
{
iter = gt_feature_node_iterator_new(cur_node);
if (iter == NULL)
return;
while ((next_node = gt_feature_node_iterator_next(iter)) && !gt_feature_node_has_type(next_node, feature_type_gene));
gt_feature_node_iterator_delete(iter);
if (NULL == (cur_node = next_node))
return 0;
}
if(!gt_feature_node_has_type(cur_node, feature_type_gene))
return 0;
// find name of gene
gene_name = gt_feature_node_get_attribute(cur_node, "Name");
if (gene_name == NULL)
return;
if ( 1 != sscanf(gt_str_get(gt_genome_node_get_seqid(cur_node)), "Chr%d", &chr_num))
return 0;
TSS = (gt_feature_node_get_strand(cur_node) == GT_STRAND_FORWARD) ? gt_genome_node_get_start(cur_node) : gt_genome_node_get_end(cur_node);
// now figure out the overlapping gene
if (! (overlap_name = CpGIOverlap_stream_find_gene_overlap( context, TSS, chr_num)))
return 0;
// save the score into the node
gt_feature_node_set_attribute(cur_node, "cpgi_at_tss", overlap_name);
return 0;
}
}
return err_num;
}
static int CpGI_score_stream_next(GtNodeStream * ns,
GtGenomeNode ** gn,
GtError * err)
{
GtGenomeNode * cur_node;
int err_num = 0;
*gn = NULL;
CpGI_score_stream * score_stream;
unsigned long island_start;
unsigned long island_end;
float island_score;
int chromosome_num;
GtStr * seqID_gtstr;
char * seqID_str;
char * num_cg_str;
unsigned long num_cg = 0;
score_stream = CpGI_score_stream_cast(ns);
// find the CpGI's, process methylome score
if(!gt_node_stream_next(score_stream->in_stream,
&cur_node,
err
) && cur_node != NULL
)
{
*gn = cur_node;
// try casting as a feature node so we can test type
if(!gt_genome_node_try_cast(gt_feature_node_class(), cur_node))
{
return 0;
}
else // we found a feature node
{
if(!gt_feature_node_has_type(cur_node, feature_type_CpGI))
return 0;
#if DEBUG_SCORE
printf("found CpGI\n");
#endif
island_start = gt_genome_node_get_start(cur_node);
island_end = gt_genome_node_get_end(cur_node);
seqID_gtstr = gt_genome_node_get_seqid(cur_node);
seqID_str = gt_str_get(seqID_gtstr);
sscanf(seqID_str, "Chr%d", &chromosome_num);
num_cg_str = gt_feature_node_get_attribute(cur_node, "sumcg");
if (!num_cg_str)
return 0;
sscanf(num_cg_str, "%d", &num_cg);
// now figure out the score
island_score = CpGI_score_stream_score_island(score_stream ,
chromosome_num,
num_cg,
island_start,
island_end);
// gt_str_delete(seqID_gtstr);
// save the score into the node
gt_feature_node_set_score(cur_node, island_score);
return 0;
}
}
return err_num;
}