void sbbs_t::create_sif_dat(char *siffile, char *datfile)
{
char *buf;
int file;
if((buf=(char *)malloc(SIF_MAXBUF))==NULL) {
errormsg(WHERE,ERR_ALLOC,siffile,SIF_MAXBUF);
return;
}
memset(buf,SIF_MAXBUF,0); /* initialize to null */
sif(siffile,buf,SIF_MAXBUF);
if((file=nopen(datfile,O_WRONLY|O_TRUNC|O_CREAT))==-1) {
free(buf);
errormsg(WHERE,ERR_OPEN,datfile,O_WRONLY|O_TRUNC|O_CREAT);
return;
}
write(file,buf,strlen(buf));
close(file);
free(buf);
}
// initial step:
// 0) id all indels in full_pr
// 1) convert these into genotype elements (including simple homopolymer normalization)
// 2) reduce to two most popular alleles for all alleles which conflict
// 3) print lots of stuff out for the active region
//
void
starling_pos_processor_base::
get_region_haplotypes(const known_pos_range full_pr,
const known_pos_range /*active_pr*/) {
// only works for the first sample right now:
static const unsigned sample_no(0);
//for(unsigned sample_no(0);sample_no<_n_samples;++sample_no)..
sample_info& sif(sample(sample_no));
// step 0: identify all candidate indels in full region:
htype_buffer hdata;
// iterate through indels in each region, find and iterate through supporting reads for each indel:
//
const indel_buffer& ibuff(sif.indel_sync().ibuff());
typedef indel_buffer::const_iterator ciiter;
const std::pair<ciiter,ciiter> ipair(ibuff.pos_range_iter(full_pr.begin_pos,full_pr.end_pos));
for (ciiter i(ipair.first); i!=ipair.second; ++i) {
const indel_key& ik(i->first);
const indel_data& id(get_indel_data(i));
// check that indel really falls into candidate range:
if (! (full_pr.is_pos_intersect(ik.pos) ||
full_pr.is_pos_intersect(ik.right_pos()))) continue;
// only consider candidates:
if (! sif.indel_sync().is_candidate_indel(_client_opt,ik,id)) continue;
get_htypes_for_indel(_client_dopt,sif.read_buff,_ref,ik,id,hdata);
// convert indel to htype_element and insert:
// convert_indel_to_htype(ik,id,_ref_seq);
}
#ifdef DEBUG_HTYPE
if (! hdata.empty()) hdata.dump(std::cerr);
#endif
#if 0
// get expanded region for read search:
const known_pos_range expanded_pr(full_pr.begin_pos-(_client_opt.max_indel_size+XXXMAX_READ_SIZE),full_pr.end_pos);
// iterate through reads in region:
for (pos_t pos(begin); pos<end; ++pos) {
read_segment_iter ri(sif.read_buff.get_pos_read_segment_iter(pos));
for (read_segment_iter::ret_val r; true; ri.next()) {
r=ri.get_ptr();
if (NULL==r.first) break;
read_segment& rseg(r.first->get_segment(r.second));
{
indel_set_t cal_indels;
get_alignment_indels(cal,opt.max_indel_size,cal_indels);
}
//rseg.blah....
}
}
#endif
#if 0
const indel_buffer& ibuff(sif.indel_sync().ibuff());
const std::pair<ciiter,ciiter> ipair(ibuff.pos_range_iter(full_pr.begin_pos,full_pr.end_pos));
for (ciiter i(ipair.first); i!=ipair.second; ++i) {
const indel_key& ik(i->first);
const indel_data& id(get_indel_data(i));
// check that indel really falls into candidate range:
if (! (full_pr.is_pos_intersect(ik.pos()) ||
full_pr.is_pos_intersect(ik.right_pos()))) continue;
// only consider candidates:
if (! sif.indel_sync().is_candidate_indel(_client_opt,ik,id)) continue;
// convert indel to htype_element and insert:
convert_indel_to_htype(ik,id,_ref_seq);
}
#endif
#if 0
const known_pos_range full_pr(start_pos,start_pos+region_size);
const known_pos_range active_pr(full_start_pos+region_size/4,start_pos+region_size);
const pos_t begin(region_size/4);
const pos_t end(begin+(region_size/2));
// step 1: identify loci in region:
const indel_buffer& ibuff(isync.ibuff());
typedef indel_buffer::const_iterator ciiter;
const std::pair<ciiter,ciiter> ipair(ibuff.pos_range_iter(pr.begin_pos,pr.end_pos));
for (ciiter i(ipair.first); i!=ipair.second; ++i) {
const indel_key& ik(i->first);
const indel_data& id(get_indel_data(i));
#if 1
// iterate through reads in region:
for (pos_t pos(begin); pos<end; ++pos) {
read_segment_iter ri(sif.read_buff.get_pos_read_segment_iter(pos));
for (read_segment_iter::ret_val r; true; ri.next()) {
r=ri.get_ptr();
if (NULL==r.first) break;
read_segment& rseg(r.first->get_segment(r.second));
//rseg.blah....
}
}
#endif
}
#endif
}
