void AlleleIdentity::CalculateWindowForVariant(LocalReferenceContext seq_context, const string &local_contig_sequence, int DEBUG) {
// If we have an invalid vcf candidate, set a length zero window and exit
if (!seq_context.context_detected) {
start_window = seq_context.position0;
end_window = seq_context.position0;
return;
}
// Check for MNRs first, for InDelLengths 2,3,4
if (status.isIndel and !status.isHPIndel and inDelLength < 5)
for (int rep_period = 2; rep_period < 5; rep_period++)
if (IdentifyMultiNucRepeatSection(local_contig_sequence, seq_context, rep_period)) {
if (DEBUG > 0) {
cout << "MNR found in allele " << seq_context.reference_allele << " -> " << altAllele << endl;
cout << "Window for allele " << altAllele << ": (" << start_window << ") ";
for (int p_idx = start_window; p_idx < end_window; p_idx++)
cout << local_contig_sequence.at(p_idx);
cout << " (" << end_window << ") " << endl;
}
return; // Found a matching period and computed window
}
// OK, not an MNR. Moving on along to InDels.
// need at least one anchor base left and right of InDel allele
if (status.isIndel) {
if (status.isDeletion) {
start_window = seq_context.my_hp_start_pos.at(anchor_length) - 1;
end_window = seq_context.right_hp_start;
}
else { // Insertions require a bit more thought
if (altAllele.at(anchor_length) == altAllele.at(anchor_length - 1))
start_window = seq_context.my_hp_start_pos.at(anchor_length - 1) - 1;
else
start_window = seq_context.position0 + anchor_length - 1; // 1 anchor base before we insert a new HP
if (start_window < 0)
start_window = 0; // Safety for something happening in the first HP of the ref. and not left-aligned...
end_window = seq_context.right_hp_start;
if (altAllele.at(altAllele.length() - 1) == seq_context.ref_right_hp_base)
end_window += seq_context.right_hp_length;
}
if ((unsigned int)end_window < local_contig_sequence.length())
end_window++; // anchor base to the right
}
else {
// SNPs and MNVs are 1->1 base replacements
// make window as short as possible, only around bases to be replaced
// Think: include full HPs affected like in the InDel case? <- no for now, like in old code
start_window = seq_context.position0;
end_window = seq_context.position0 + seq_context.reference_allele.length();
} // */
if (DEBUG > 0) {
cout << "Window for allele " << altAllele << ": (" << start_window << ") ";
for (int p_idx = start_window; p_idx < end_window; p_idx++)
cout << local_contig_sequence.at(p_idx);
cout << " (" << end_window << ") " << endl;
}
}
void AlleleIdentity::CalculateWindowForVariant(const LocalReferenceContext &seq_context, int DEBUG,
const ReferenceReader &ref_reader, int chr_idx) {
// If we have an invalid vcf candidate, set a length zero window and exit
if (!seq_context.context_detected or status.isProblematicAllele) {
start_window = seq_context.position0;
end_window = seq_context.position0;
return;
}
// Check for MNRs first, for InDelLengths 2,3,4,5
if (status.isIndel and !status.isHPIndel and inDelLength < 5)
for (int rep_period = 2; rep_period < 6; rep_period++)
if (IdentifyMultiNucRepeatSection(seq_context, rep_period, ref_reader, chr_idx)) {
if (DEBUG > 0) {
cout << "MNR found in allele " << seq_context.reference_allele << " -> " << altAllele << endl;
cout << "Window for allele " << altAllele << ": (" << start_window << ") ";
for (int p_idx = start_window; p_idx < end_window; p_idx++)
cout << ref_reader.base(chr_idx,p_idx);
cout << " (" << end_window << ") " << endl;
}
return; // Found a matching period and computed window
}
// not an MNR. Moving on along to InDels.
if (status.isIndel) {
// Default variant window
end_window = seq_context.right_hp_start +1; // Anchor base to the right of allele
start_window = seq_context.position0;
// Adjustments if necessary
if (status.isDeletion)
if (seq_context.my_hp_start_pos[left_anchor] == seq_context.my_hp_start_pos[0])
start_window = seq_context.my_hp_start_pos[0] - 1;
if (status.isInsertion) {
if (left_anchor == 0) {
start_window = seq_context.my_hp_start_pos[0] - 1;
}
else if (altAllele[left_anchor] == altAllele[left_anchor - 1] and
seq_context.position0 > (seq_context.my_hp_start_pos[left_anchor - 1] - 1)) {
start_window = seq_context.my_hp_start_pos[left_anchor - 1] - 1;
}
if (altAllele[altAllele.length() - 1] == seq_context.ref_right_hp_base) {
end_window += seq_context.right_hp_length;
}
}
// Safety
if (start_window < 0)
start_window = 0;
if (end_window > ref_reader.chr_size(chr_idx))
end_window = ref_reader.chr_size(chr_idx);
}
else {
// SNPs and MNVs are 1->1 base replacements
start_window = seq_context.position0;
end_window = seq_context.position0 + seq_context.reference_allele.length();
} // */
if (DEBUG > 0) {
cout << "Window for allele " << altAllele << ": (" << start_window << ") ";
for (int p_idx = start_window; p_idx < end_window; p_idx++)
cout << ref_reader.base(chr_idx,p_idx);
cout << " (" << end_window << ") " << endl;
}
}