// Calculate the expected ratio of reads observed to support each
// allele. Note that for sites and single breakpoints this is expected
// to match the sample allele ratio, however for indels this can
// change as a funciton of indel and read length.
//
// Note this routine does not accoung for overlapping indels
//
static
void
get_het_observed_allele_ratio(const unsigned read_length,
const unsigned min_overlap,
const indel_key& ik,
const double het_allele_ratio,
double& log_ref_prob,
double& log_indel_prob) {
assert((ik.type==INDEL::INSERT) ||
(ik.type==INDEL::DELETE) ||
(ik.type == INDEL::SWAP));
// the expected relative read depth for two breakpoints separated by a distance of 0:
const unsigned base_expect( (read_length+1)<(2*min_overlap) ? 0 : (read_length+1)-(2*min_overlap) );
// Get expected relative read depth for the shorter and longer
// paths of a general sequence replacement. Note this includes
// basic insertions and deletions, in these cases
// spath_break_distance is 0 and spath_expect equals base_expect:
//
const double ref_path_expect(base_expect+std::min(ik.delete_length(),base_expect));
const double indel_path_expect(base_expect+std::min(ik.insert_length(),base_expect));
const double ref_path_term((1-het_allele_ratio)*ref_path_expect);
const double indel_path_term(het_allele_ratio*indel_path_expect);
const double total_path_term(ref_path_term+indel_path_term);
if (total_path_term>0) {
const double indel_prob(indel_path_term/total_path_term);
log_ref_prob=std::log(1.-indel_prob);
log_indel_prob=std::log(indel_prob);
}
}
/// get the indel cigar and ref and indel strings used in the indel
/// summary line output
///
static
void
get_vcf_summary_strings(const indel_key& ik,
const indel_data& id,
const reference_contig_segment& ref,
std::string& vcf_indel_seq,
std::string& vcf_ref_seq) {
if (ik.is_breakpoint()) {
if (ik.type == INDEL::BP_LEFT) {
copy_ref_subseq(ref,ik.pos-1,ik.pos,vcf_ref_seq);
vcf_indel_seq = vcf_ref_seq + id.get_insert_seq() + '.';
} else if(ik.type == INDEL::BP_RIGHT) {
copy_ref_subseq(ref,ik.pos,ik.pos+1,vcf_ref_seq);
vcf_indel_seq = '.' + id.get_insert_seq() + vcf_ref_seq;
} else {
assert(0);
}
} else {
copy_ref_subseq(ref,ik.pos-1,ik.pos+ik.delete_length(),vcf_ref_seq);
copy_ref_subseq(ref,ik.pos-1,ik.pos,vcf_indel_seq);
vcf_indel_seq += id.get_insert_seq();
}
}