/**
* Gets a sorted string representation of the alleles of a variant.
*/
void bcf_alleles2string_sorted(bcf_hdr_t *h, bcf1_t *v, kstring_t *var)
{
bcf_unpack(v, BCF_UN_STR);
var->l = 0;
if (v->n_allele==2)
{
kputs(bcf_get_alt(v, 0), var);
kputc(',', var);
kputs(bcf_get_alt(v, 1), var);
}
else
{
char** allele = bcf_get_allele(v);
char** temp = (char**) malloc((bcf_get_n_allele(v)-1)*sizeof(char*));
for (int32_t i=1; i<v->n_allele; ++i)
{
temp[i-1] = allele[i];
}
std::qsort(temp, bcf_get_n_allele(v)-1, sizeof(char*), cmpstr);
kputs(bcf_get_alt(v, 0), var);
for (int32_t i=0; i<v->n_allele-1; ++i)
{
kputc(',', var);
kputs(temp[i], var);
}
free(temp);
}
}
/**
* Extract reference sequence region for motif discovery in a fuzzy fashion.
*/
void CandidateRegionExtractor::extract_regions_by_fuzzy_alignment(bcf_hdr_t* h, bcf1_t* v, Variant& variant)
{
if (debug)
{
if (debug) std::cerr << "********************************************\n";
std::cerr << "EXTRACTIING REGION BY FUZZY ALIGNMENT\n\n";
}
VNTR& vntr = variant.vntr;
const char* chrom = bcf_get_chrom(h, v);
int32_t min_beg1 = bcf_get_pos1(v);
int32_t max_end1 = min_beg1;
//merge candidate search region
for (size_t i=1; i<bcf_get_n_allele(v); ++i)
{
std::string ref(bcf_get_alt(v, 0));
std::string alt(bcf_get_alt(v, i));
int32_t pos1 = bcf_get_pos1(v);
trim(pos1, ref, alt);
if (debug)
{
std::cerr << "indel fragment : " << (ref.size()<alt.size()? alt : ref) << "\n";
std::cerr << " : " << ref << ":" << alt << "\n";
}
min_beg1 = fuzzy_left_align(chrom, pos1, ref, alt, 3);
max_end1 = fuzzy_right_align(chrom, pos1 + ref.size() - 1, ref, alt, 3);
int32_t seq_len;
char* seq = faidx_fetch_seq(fai, chrom, min_beg1-1, max_end1-1, &seq_len);
if (debug)
{
std::cerr << "FUZZY REGION " << min_beg1 << "-" << max_end1 << " (" << max_end1-min_beg1+1 <<") " << "\n";
std::cerr << " " << seq << "\n";
}
if (seq_len) free(seq);
}
int32_t seq_len;
char* seq = faidx_fetch_seq(fai, chrom, min_beg1-1, max_end1-1, &seq_len);
if (debug)
{
std::cerr << "FINAL FUZZY REGION " << min_beg1 << "-" << max_end1 << " (" << max_end1-min_beg1+1 <<") " << "\n";
std::cerr << " " << seq << "\n";
}
vntr.exact_repeat_tract = seq;
vntr.exact_rbeg1 = min_beg1;
if (seq_len) free(seq);
}
/**
* Checks if a vntr is a homopolymer.
*/
bool CandidateRegionExtractor::is_homopolymer(bcf_hdr_t* h, bcf1_t* v)
{
bool is_homopolymer = false;
uint32_t ref_len = strlen(bcf_get_ref(v));
for (size_t i=1; i<bcf_get_n_allele(v); ++i)
{
std::string ref(bcf_get_alt(v, 0));
std::string alt(bcf_get_alt(v, i));
int32_t pos1 = bcf_get_pos1(v);
}
return is_homopolymer;
}
/**
* Gets a string representation of a variant.
*/
void bcf_variant2string(bcf_hdr_t *h, bcf1_t *v, kstring_t *var)
{
bcf_unpack(v, BCF_UN_STR);
var->l = 0;
kputs(bcf_get_chrom(h, v), var);
kputc(':', var);
kputw(bcf_get_pos1(v), var);
kputc(':', var);
for (int32_t i=0; i<v->n_allele; ++i)
{
if (i) kputc(',', var);
kputs(bcf_get_alt(v, i), var);
}
}
/**
* Gets a string representation of the variant.
*/
std::string Variant::get_variant_string()
{
kstring_t var = {0,0,0};
bcf_unpack(v, BCF_UN_STR);
var.l = 0;
kputs(bcf_get_chrom(h, v), &var);
kputc(':', &var);
kputw(bcf_get_pos1(v), &var);
kputc(':', &var);
for (size_t i=0; i<bcf_get_n_allele(v); ++i)
{
if (i) kputc('/', &var);
kputs(bcf_get_alt(v, i), &var);
}
std::string str(var.s);
if (var.m) free(var.s);
return str;
}
/**
* Extract reference sequence region for motif discovery.
*
* The input is a VCF record that contains an indel.
*
* If the the indel has multiple alleles, it will examine all
* alleles.
*
* todo: is might be a good idea to combine this step with motif detection
* since there seems to be a need to have an iterative process here
* to ensure a good candidate motif is chosen. *
*/
void CandidateRegionExtractor::extract_regions_by_exact_alignment(bcf_hdr_t* h, bcf1_t* v, Variant& variant)
{
if (debug)
{
if (debug) std::cerr << "********************************************\n";
std::cerr << "EXTRACTIING REGION BY EXACT LEFT AND RIGHT ALIGNMENT\n\n";
}
VNTR& vntr = variant.vntr;
const char* chrom = bcf_get_chrom(h, v);
int32_t min_beg1 = bcf_get_pos1(v);
int32_t max_end1 = min_beg1;
if (debug)
{
bcf_print_liten(h, v);
}
//merge candidate search region
for (size_t i=1; i<bcf_get_n_allele(v); ++i)
{
std::string ref(bcf_get_alt(v, 0));
std::string alt(bcf_get_alt(v, i));
int32_t pos1 = bcf_get_pos1(v);
//this prevents introduction of flanks that do not harbour the repeat unit
trim(pos1, ref, alt);
int32_t end1 = pos1 + ref.size() - 1;
right_align(chrom, end1, ref, alt);
int32_t beg1 = end1 - ref.size() + 1;
left_align(chrom, beg1, ref, alt);
min_beg1 = beg1<min_beg1 ? beg1 : min_beg1;
max_end1 = end1>max_end1 ? end1 : max_end1;
int32_t seq_len;
char* seq = faidx_fetch_seq(fai, chrom, min_beg1-1, max_end1-1, &seq_len);
if (debug)
{
std::cerr << "EXACT REGION " << min_beg1 << "-" << max_end1 << " (" << max_end1-min_beg1+1 <<") from " << pos1 << ":" << ref << ":" << alt << "\n";
std::cerr << " " << seq << "\n";
}
if (seq_len) free(seq);
}
int32_t seq_len;
char* seq = faidx_fetch_seq(fai, chrom, min_beg1-1, max_end1-1, &seq_len);
if (debug)
{
std::cerr << "FINAL EXACT REGION " << min_beg1 << "-" << max_end1 << " (" << max_end1-min_beg1+1 <<") " << "\n";
std::cerr << " " << seq << "\n";
}
vntr.exact_repeat_tract = seq;
vntr.rid = bcf_get_rid(v);
vntr.exact_rbeg1 = min_beg1;
vntr.exact_rend1 = max_end1;
if (seq_len) free(seq);
}