Alignment strip_from_start(const Alignment& aln, size_t drop) { if (!drop) return aln; Alignment res; res.set_name(aln.name()); res.set_score(aln.score()); //cerr << "drop " << drop << " from start" << endl; res.set_sequence(aln.sequence().substr(drop)); if (!aln.has_path()) return res; *res.mutable_path() = cut_path(aln.path(), drop).second; assert(res.has_path()); if (alignment_to_length(res) != res.sequence().size()) { cerr << "failed!!! drop from start 轰" << endl; cerr << pb2json(res) << endl << endl; assert(false); } return res; }
Alignment strip_from_end(const Alignment& aln, size_t drop) { if (!drop) return aln; Alignment res; res.set_name(aln.name()); res.set_score(aln.score()); //cerr << "drop " << drop << " from end" << endl; size_t cut_at = aln.sequence().size()-drop; //cerr << "Cut at " << cut_at << endl; res.set_sequence(aln.sequence().substr(0, cut_at)); if (!aln.has_path()) return res; *res.mutable_path() = cut_path(aln.path(), cut_at).first; assert(res.has_path()); if (alignment_to_length(res) != res.sequence().size()) { cerr << "failed!!! drop from end 轰" << endl; cerr << pb2json(res) << endl << endl; assert(false); } return res; }
void Aligner::gssw_mapping_to_alignment(gssw_graph* graph, gssw_graph_mapping* gm, Alignment& alignment, bool print_score_matrices) { alignment.clear_path(); alignment.set_score(gm->score); alignment.set_query_position(0); Path* path = alignment.mutable_path(); //alignment.set_cigar(graph_cigar(gm)); gssw_graph_cigar* gc = &gm->cigar; gssw_node_cigar* nc = gc->elements; int to_pos = 0; int from_pos = gm->position; //cerr << "gm->position " << gm->position << endl; string& to_seq = *alignment.mutable_sequence(); //cerr << "-------------" << endl; if (print_score_matrices) { gssw_graph_print_score_matrices(graph, to_seq.c_str(), to_seq.size(), stderr); //cerr << alignment.DebugString() << endl; } for (int i = 0; i < gc->length; ++i, ++nc) { if (i > 0) from_pos = 0; // reset for each node after the first // check that the current alignment has a non-zero length gssw_cigar* c = nc->cigar; int l = c->length; if (l == 0) continue; gssw_cigar_element* e = c->elements; Node* from_node = (Node*) nc->node->data; string& from_seq = *from_node->mutable_sequence(); Mapping* mapping = path->add_mapping(); mapping->mutable_position()->set_node_id(nc->node->id); mapping->mutable_position()->set_offset(from_pos); mapping->set_rank(path->mapping_size()); //cerr << from_node->id() << ":" << endl; for (int j=0; j < l; ++j, ++e) { Edit* edit; int32_t length = e->length; //cerr << e->length << e->type << endl; switch (e->type) { case 'M': case 'X': case 'N': { // do the sequences match? // emit a stream of "SNPs" and matches int h = from_pos; int last_start = from_pos; int k = to_pos; for ( ; h < from_pos + length; ++h, ++k) { //cerr << h << ":" << k << " " << from_seq[h] << " " << to_seq[k] << endl; if (from_seq[h] != to_seq[k]) { // emit the last "match" region if (h-last_start > 0) { edit = mapping->add_edit(); edit->set_from_length(h-last_start); edit->set_to_length(h-last_start); } // set up the SNP edit = mapping->add_edit(); edit->set_from_length(1); edit->set_to_length(1); edit->set_sequence(to_seq.substr(k,1)); last_start = h+1; } } // handles the match at the end or the case of no SNP if (h-last_start > 0) { edit = mapping->add_edit(); edit->set_from_length(h-last_start); edit->set_to_length(h-last_start); } to_pos += length; from_pos += length; } break; case 'D': edit = mapping->add_edit(); edit->set_from_length(length); edit->set_to_length(0); from_pos += length; break; case 'I': edit = mapping->add_edit(); edit->set_from_length(0); edit->set_to_length(length); edit->set_sequence(to_seq.substr(to_pos, length)); to_pos += length; break; case 'S': // note that soft clips and insertions are semantically equivalent // and can only be differentiated by their position in the read // with soft clips coming at the start or end edit = mapping->add_edit(); edit->set_from_length(0); edit->set_to_length(length); edit->set_sequence(to_seq.substr(to_pos, length)); to_pos += length; break; default: cerr << "error:[Aligner::gssw_mapping_to_alignment] " << "unsupported cigar op type " << e->type << endl; exit(1); break; } } //cerr << "path to_length " << path_to_length(*path) << endl; } // set identity alignment.set_identity(identity(alignment.path())); }