1 // This file is part of the Mussa source distribution.
2 // http://mussa.caltech.edu/
3 // Contact author: Tristan De Buysscher, tristan@caltech.edu
5 // This program and all associated source code files are Copyright (C) 2005
6 // the California Institute of Technology, Pasadena, CA, 91125 USA. It is
7 // under the GNU Public License; please see the included LICENSE.txt
8 // file for more information, or contact Tristan directly.
11 // This file is part of the Mussa source distribution.
12 // http://mussa.caltech.edu/
13 // Contact author: Tristan De Buysscher, tristan@caltech.edu
15 // This program and all associated source code files are Copyright (C) 2005
16 // the California Institute of Technology, Pasadena, CA, 91125 USA. It is
17 // under the GNU Public License; please see the included LICENSE.txt
18 // file for more information, or contact Tristan directly.
21 // ----------------------------------------
22 // ---------- sequence.cc -----------
23 // ----------------------------------------
24 #include <boost/filesystem/fstream.hpp>
25 namespace fs = boost::filesystem;
27 #include <boost/spirit/core.hpp>
28 #include <boost/spirit/actor/push_back_actor.hpp>
29 #include <boost/spirit/iterator/file_iterator.hpp>
30 #include <boost/spirit/utility/chset.hpp>
31 namespace spirit = boost::spirit;
33 #include "alg/sequence.hpp"
34 #include "mussa_exceptions.hpp"
50 annot::annot(int begin, int end, std::string type, std::string name)
62 bool operator==(const annot& left, const annot& right)
64 return ((left.begin== right.begin) and
65 (left.end == right.end) and
66 (left.type == right.type) and
67 (left.name == right.name));
70 motif::motif(int begin, std::string motif)
71 : annot(begin, begin+motif.size(), "motif", motif),
81 Sequence::Sequence(alphabet_ref alphabet_)
94 Sequence::Sequence(const char *seq, alphabet_ref alphabet_)
99 strand(UnknownStrand),
103 set_filtered_sequence(seq, alphabet);
106 Sequence::Sequence(const std::string& seq, alphabet_ref alphabet_)
111 strand(UnknownStrand),
115 set_filtered_sequence(seq, alphabet);
118 Sequence::Sequence(const Sequence& o)
121 alphabet(o.alphabet),
122 seq_start(o.seq_start),
123 seq_count(o.seq_count),
128 motif_list(o.motif_list)
132 Sequence &Sequence::operator=(const Sequence& s)
137 alphabet = s.alphabet;
138 seq_start = s.seq_start;
139 seq_count = s.seq_count;
144 motif_list = s.motif_list;
149 static void multiplatform_getline(std::istream& in, std::string& line)
154 while(in.good() and !(c == '\012' or c == '\015') ) {
158 // if we have cr-lf eat it
160 if (c=='\012' or c == '\015') {
165 void Sequence::load_fasta(fs::path file_path, int seq_num, int start_index, int end_index)
167 load_fasta(file_path, alphabet, seq_num, start_index, end_index);
170 //! load a fasta file into a sequence
171 void Sequence::load_fasta(fs::path file_path, alphabet_ref a,
172 int seq_num, int start_index, int end_index)
174 fs::fstream data_file;
175 data_file.open(file_path, std::ios::in);
177 if (!data_file.good())
179 throw mussa_load_error("Sequence File: "+file_path.string()+" not found");
182 load_fasta(data_file, a, seq_num, start_index, end_index);
183 } catch(sequence_empty_error e) {
184 // there doesn't appear to be any sequence
185 // catch and rethrow to include the filename
186 std::stringstream msg;
187 msg << "The selected sequence in "
188 << file_path.native_file_string()
189 << " appears to be empty";
190 throw sequence_empty_error(msg.str());
191 } catch(sequence_empty_file_error e) {
192 std::stringstream errormsg;
193 errormsg << file_path.native_file_string()
194 << " did not have any fasta sequences" << std::endl;
195 throw sequence_empty_file_error(errormsg.str());
196 } catch(sequence_invalid_load_error e) {
197 std::ostringstream msg;
198 msg << file_path.native_file_string();
199 msg << " " << e.what();
200 throw sequence_invalid_load_error(msg.str());
205 void Sequence::load_fasta(std::istream& file,
206 int seq_num, int start_index, int end_index)
208 load_fasta(file, alphabet, seq_num, start_index, end_index);
212 Sequence::load_fasta(std::istream& data_file, alphabet_ref a,
214 int start_index, int end_index)
216 std::string file_data_line;
217 int header_counter = 0;
218 size_t line_counter = 0;
219 bool read_seq = true;
220 std::string rev_comp;
221 std::string sequence_raw;
222 std::string seq_tmp; // holds sequence during basic filtering
223 const Alphabet &alpha = get_alphabet(a);
226 throw mussa_load_error("fasta sequence number is 1 based (can't be 0)");
229 // search for the header of the fasta sequence we want
230 while ( (!data_file.eof()) && (header_counter < seq_num) )
232 multiplatform_getline(data_file, file_data_line);
234 if (file_data_line.substr(0,1) == ">")
238 if (header_counter > 0) {
239 header = file_data_line.substr(1);
243 while ( !data_file.eof() && read_seq ) {
244 multiplatform_getline(data_file,file_data_line);
246 if (file_data_line.substr(0,1) == ">")
249 for (std::string::const_iterator line_i = file_data_line.begin();
250 line_i != file_data_line.end();
253 if(alpha.exists(*line_i)) {
254 sequence_raw += *line_i;
256 std::ostringstream msg;
257 msg << "Unrecognized characters in fasta sequence at line ";
259 throw sequence_invalid_load_error(msg.str());
265 // Lastly, if subselection of the sequence was specified we keep cut out
266 // and only keep that part
267 // end_index = 0 means no end was specified, so cut to the end
269 end_index = sequence_raw.size();
271 // sequence filtering for upcasing agctn and convert non AGCTN to N
272 if (end_index-start_index <= 0) {
273 std::string msg("The selected sequence appears to be empty");
274 throw sequence_empty_error(msg);
276 set_filtered_sequence(sequence_raw, a, start_index, end_index-start_index);
278 std::string errormsg("There were no fasta sequences");
279 throw sequence_empty_file_error(errormsg);
283 void Sequence::set_filtered_sequence(const std::string &in_seq,
284 alphabet_ref alphabet_,
289 alphabet = alphabet_;
291 count = in_seq.size() - start;
292 boost::shared_ptr<seq_string> new_seq(new seq_string);
293 new_seq->reserve(count);
295 // finally, the actual conversion loop
296 const Alphabet& alpha_impl = get_alphabet(); // go get one of our actual alphabets
297 std::string::const_iterator seq_i = in_seq.begin()+start;
298 for(size_type i = 0; i != count; ++i, ++seq_i)
300 if (alpha_impl.exists(*seq_i)) {
301 new_seq->append(1, toupper(*seq_i));
303 new_seq->append(1, 'N');
314 Sequence::load_annot(fs::path file_path, int start_index, int end_index)
316 fs::fstream data_stream(file_path, std::ios::in);
319 throw mussa_load_error("Sequence File: " + file_path.string() + " not found");
322 // so i should probably be passing the parse function some iterators
323 // but the annotations files are (currently) small, so i think i can
324 // get away with loading the whole file into memory
327 while(data_stream.good()) {
334 parse_annot(data, start_index, end_index);
335 } catch(annotation_load_error e) {
336 std::ostringstream msg;
337 msg << file_path.native_file_string()
340 throw annotation_load_error(msg.str());
344 /* If this works, yikes, this is some brain hurting code.
346 * what's going on is that when pb_annot is instantiated it stores references
347 * to begin, end, name, type, declared in the parse function, then
348 * when operator() is called it grabs values from those references
349 * and uses that to instantiate an annot object and append that to our
352 * This weirdness is because the spirit library requires that actions
353 * conform to a specific prototype operator()(IteratorT, IteratorT)
354 * which doesn't provide any useful opportunity for me to actually
355 * grab the results of our parsing.
357 * so I instantiate this structure in order to have a place to grab
361 struct push_back_annot {
362 std::list<annot>& annot_list;
369 push_back_annot(std::list<annot>& annot_list_,
375 : annot_list(annot_list_),
384 void operator()(std::string::const_iterator,
385 std::string::const_iterator) const
387 //std::cout << "adding annot: " << begin << "|" << end << "|" << name << "|" << type << std::endl;
388 annot_list.push_back(annot(begin, end, name, type));
393 struct push_back_seq {
394 std::list<Sequence>& seq_list;
399 push_back_seq(std::list<Sequence>& seq_list_,
403 : seq_list(seq_list_),
410 void operator()(std::string::const_iterator,
411 std::string::const_iterator) const
413 // filter out newlines from our sequence
415 for(std::string::const_iterator seq_i = seq.begin();
419 if (*seq_i != '\015' && *seq_i != '\012') new_seq += *seq_i;
421 //std::cout << "adding seq: " << name << " " << new_seq << std::endl;
424 s.set_fasta_header(name);
425 seq_list.push_back(s);
431 Sequence::parse_annot(std::string data, int start_index, int end_index)
438 std::list<annot> parsed_annots;
439 std::list<Sequence> query_seqs;
442 bool ok = spirit::parse(data.begin(), data.end(),
449 )[spirit::assign_a(species)] >>
453 ( // ignore html tags
454 *(spirit::space_p) >>
456 +(~spirit::ch_p('>')) >>
461 ( // parse an absolute location name
462 (spirit::uint_p[spirit::assign_a(start)] >>
464 spirit::uint_p[spirit::assign_a(end)] >>
469 )[spirit::assign_a(name)] >>
476 )[spirit::assign_a(type)]
478 // to understand how this group gets set
479 // read the comment above struct push_back_annot
480 )[push_back_annot(parsed_annots, start, end, type, name, parsed)]
482 ((spirit::ch_p('>')|spirit::str_p(">")) >>
483 (*(spirit::print_p))[spirit::assign_a(name)] >>
485 (+(spirit::chset<>(Alphabet::nucleic_cstr)))[spirit::assign_a(seq)]
486 )[push_back_seq(query_seqs, name, seq, parsed)]
493 std::stringstream msg;
494 msg << "Error parsing annotation #" << parsed;
495 throw annotation_load_error(msg.str());
497 // add newly parsed annotations to our sequence
498 std::copy(parsed_annots.begin(), parsed_annots.end(), std::back_inserter(annots));
499 // go seearch for query sequences
500 find_sequences(query_seqs.begin(), query_seqs.end());
503 void Sequence::add_annotation(const annot& a)
508 const std::list<annot>& Sequence::annotations() const
514 Sequence::subseq(int start, int count)
521 // there might be an off by one error with start+count > size()
522 if ( count == npos || start+count > size()) {
523 count = size()-start;
525 Sequence new_seq(*this);
526 new_seq.parent = this;
527 new_seq.seq_start = seq_start+start;
528 new_seq.seq_count = count;
530 new_seq.motif_list = motif_list;
531 new_seq.annots.clear();
532 // attempt to copy & reannotate position based annotations
533 int end = start+count;
535 for(std::list<annot>::const_iterator annot_i = annots.begin();
536 annot_i != annots.end();
539 int annot_begin= annot_i->begin;
540 int annot_end = annot_i->end;
542 if (annot_begin < end) {
543 if (annot_begin >= start) {
544 annot_begin -= start;
549 if (annot_end < end) {
555 annot new_annot(annot_begin, annot_end, annot_i->type, annot_i->name);
556 new_seq.annots.push_back(new_annot);
563 std::string Sequence::create_reverse_map() const
565 std::string rc_map(256, '~');
566 // if we're rna, use U instead of T
567 // we might want to add an "is_rna" to sequence at somepoint
568 char TU = (alphabet == reduced_rna_alphabet) ? 'U' : 'T';
569 char tu = (alphabet == reduced_rna_alphabet) ? 'u' : 't';
570 rc_map['A'] = TU ; rc_map['a'] = tu ;
571 rc_map['T'] = 'A'; rc_map['t'] = 'a';
572 rc_map['U'] = 'A'; rc_map['u'] = 'a';
573 rc_map['G'] = 'C'; rc_map['g'] = 'c';
574 rc_map['C'] = 'G'; rc_map['c'] = 'g';
575 rc_map['M'] = 'K'; rc_map['m'] = 'k';
576 rc_map['R'] = 'Y'; rc_map['r'] = 'y';
577 rc_map['W'] = 'W'; rc_map['w'] = 'w';
578 rc_map['S'] = 'S'; rc_map['s'] = 's';
579 rc_map['Y'] = 'R'; rc_map['y'] = 'r';
580 rc_map['K'] = 'M'; rc_map['k'] = 'm';
581 rc_map['V'] = 'B'; rc_map['v'] = 'b';
582 rc_map['H'] = 'D'; rc_map['h'] = 'd';
583 rc_map['D'] = 'H'; rc_map['d'] = 'h';
584 rc_map['B'] = 'V'; rc_map['b'] = 'v';
585 rc_map['N'] = 'N'; rc_map['n'] = 'n';
586 rc_map['X'] = 'X'; rc_map['x'] = 'x';
590 rc_map['~'] = '~'; // not really needed, but perhaps it's clearer.
594 Sequence Sequence::rev_comp() const
596 std::string rev_comp;
597 rev_comp.reserve(length());
599 std::string rc_map = create_reverse_map();
601 // reverse and convert
602 Sequence::const_reverse_iterator seq_i;
603 Sequence::const_reverse_iterator seq_end = rend();
604 for(seq_i = rbegin();
608 rev_comp.append(1, rc_map[*seq_i]);
610 return Sequence(rev_comp, alphabet);
613 void Sequence::set_fasta_header(std::string header_)
618 void Sequence::set_species(const std::string& name)
623 std::string Sequence::get_species() const
630 Sequence::get_fasta_header() const
636 Sequence::get_name() const
638 if (header.size() > 0)
640 else if (species.size() > 0)
646 const Alphabet& Sequence::get_alphabet() const
648 return get_alphabet(alphabet);
651 const Alphabet& Sequence::get_alphabet(alphabet_ref alpha) const
654 case reduced_dna_alphabet:
655 return Alphabet::reduced_dna_alphabet();
656 case reduced_rna_alphabet:
657 return Alphabet::reduced_rna_alphabet();
658 case reduced_nucleic_alphabet:
659 return Alphabet::reduced_nucleic_alphabet();
660 case nucleic_alphabet:
661 return Alphabet::nucleic_alphabet();
662 case protein_alphabet:
663 return Alphabet::protein_alphabet();
665 throw std::runtime_error("unrecognized alphabet type");
670 void Sequence::set_sequence(const std::string& s, alphabet_ref a)
672 set_filtered_sequence(s, a);
675 std::string Sequence::get_sequence() const
680 return std::string();
683 Sequence::const_reference Sequence::operator[](Sequence::size_type i) const
688 Sequence::const_reference Sequence::at(Sequence::size_type i) const
690 if (!seq) throw std::out_of_range("empty sequence");
691 return seq->at(i+seq_start);
701 strand = UnknownStrand;
708 const char *Sequence::c_str() const
711 return seq->c_str()+seq_start;
716 Sequence::const_iterator Sequence::begin() const
718 if (seq and seq_count != 0)
719 return seq->begin()+seq_start;
721 return Sequence::const_iterator(0);
724 Sequence::const_iterator Sequence::end() const
726 if (seq and seq_count != 0) {
727 return seq->begin() + seq_start + seq_count;
729 return Sequence::const_iterator(0);
733 Sequence::const_reverse_iterator Sequence::rbegin() const
735 if (seq and seq_count != 0)
736 return seq->rbegin()+(seq->size()-(seq_start+seq_count));
738 return Sequence::const_reverse_iterator();
741 Sequence::const_reverse_iterator Sequence::rend() const
743 if (seq and seq_count != 0) {
744 return rbegin() + seq_count;
746 return Sequence::const_reverse_iterator();
750 bool Sequence::empty() const
752 return (seq_count == 0) ? true : false;
755 Sequence::size_type Sequence::find_first_not_of(
756 const std::string& query,
757 Sequence::size_type index)
759 typedef std::set<std::string::value_type> sequence_set;
760 sequence_set match_set;
762 for(const_iterator query_item = query.begin();
763 query_item != query.end();
766 match_set.insert(*query_item);
768 for(const_iterator base = begin();
772 if(match_set.find(*base) == match_set.end()) {
776 return Sequence::npos;
779 Sequence::size_type Sequence::start() const
782 return seq_start - parent->start();
787 Sequence::size_type Sequence::stop() const
789 return start() + seq_count;
792 Sequence::size_type Sequence::size() const
797 Sequence::size_type Sequence::length() const
803 Sequence::save(fs::fstream &save_file)
806 std::list<annot>::iterator annots_i;
808 // not sure why, or if i'm doing something wrong, but can't seem to pass
809 // file pointers down to this method from the mussa control class
810 // so each call to save a sequence appends to the file started by mussa_class
811 //save_file.open(save_file_path.c_str(), std::ios::app);
813 save_file << "<Sequence>" << std::endl;
814 save_file << *this << std::endl;
815 save_file << "</Sequence>" << std::endl;
817 save_file << "<Annotations>" << std::endl;
818 save_file << species << std::endl;
819 for (annots_i = annots.begin(); annots_i != annots.end(); ++annots_i)
821 save_file << annots_i->begin << " " << annots_i->end << " " ;
822 save_file << annots_i->name << " " << annots_i->type << std::endl;
824 save_file << "</Annotations>" << std::endl;
829 Sequence::load_museq(fs::path load_file_path, int seq_num)
831 fs::fstream load_file;
832 std::string file_data_line;
835 std::string::size_type space_split_i;
836 std::string annot_value;
839 load_file.open(load_file_path, std::ios::in);
842 // search for the seq_num-th sequence
843 while ( (!load_file.eof()) && (seq_counter < seq_num) )
845 getline(load_file,file_data_line);
846 if (file_data_line == "<Sequence>")
849 getline(load_file, file_data_line);
850 // looks like the sequence is written as a single line
851 set_filtered_sequence(file_data_line, reduced_dna_alphabet);
852 getline(load_file, file_data_line);
853 getline(load_file, file_data_line);
854 if (file_data_line == "<Annotations>")
856 getline(load_file, file_data_line);
857 species = file_data_line;
858 while ( (!load_file.eof()) && (file_data_line != "</Annotations>") )
860 getline(load_file,file_data_line);
861 if ((file_data_line != "") && (file_data_line != "</Annotations>"))
863 // need to get 4 values...almost same code 4 times...
864 // get annot start index
865 space_split_i = file_data_line.find(" ");
866 annot_value = file_data_line.substr(0,space_split_i);
867 an_annot.begin = atoi (annot_value.c_str());
868 file_data_line = file_data_line.substr(space_split_i+1);
869 // get annot end index
870 space_split_i = file_data_line.find(" ");
871 annot_value = file_data_line.substr(0,space_split_i);
872 an_annot.end = atoi (annot_value.c_str());
874 if (space_split_i == std::string::npos) // no entry for type or name
876 std::cout << "seq, annots - no type or name\n";
880 else // else get annot type
882 file_data_line = file_data_line.substr(space_split_i+1);
883 space_split_i = file_data_line.find(" ");
884 annot_value = file_data_line.substr(0,space_split_i);
885 an_annot.type = annot_value;
886 if (space_split_i == std::string::npos) // no entry for name
888 std::cout << "seq, annots - no name\n";
891 else // get annot name
893 file_data_line = file_data_line.substr(space_split_i+1);
894 space_split_i = file_data_line.find(" ");
895 annot_value = file_data_line.substr(0,space_split_i);
896 an_annot.type = annot_value;
899 annots.push_back(an_annot); // don't forget to actually add the annot
901 //std::cout << "seq, annots: " << an_annot.start << ", " << an_annot.end
902 // << "-->" << an_annot.type << "::" << an_annot.name << std::endl;
909 void Sequence::add_motif(const Sequence& a_motif)
911 std::vector<int> motif_starts = find_motif(a_motif);
913 for(std::vector<int>::iterator motif_start_i = motif_starts.begin();
914 motif_start_i != motif_starts.end();
917 motif_list.push_back(motif(*motif_start_i, a_motif.get_sequence()));
921 void Sequence::clear_motifs()
926 const std::list<motif>& Sequence::motifs() const
932 Sequence::find_motif(const Sequence& a_motif) const
934 std::vector<int> motif_match_starts;
935 Sequence norm_motif_rc;
937 motif_match_starts.clear();
938 // std::cout << "motif is: " << norm_motif << std::endl;
940 if (a_motif.size() > 0)
942 //std::cout << "Sequence: none blank motif\n";
943 motif_scan(a_motif, &motif_match_starts);
945 norm_motif_rc = a_motif.rev_comp();;
946 // make sure not to do search again if it is a palindrome
947 if (norm_motif_rc != a_motif) {
948 motif_scan(norm_motif_rc, &motif_match_starts);
951 return motif_match_starts;
955 Sequence::motif_scan(const Sequence& a_motif, std::vector<int> * motif_match_starts) const
957 // if there's no sequence we can't scan for it?
958 // should this throw an exception?
961 std::string::size_type seq_i = 0;
962 Sequence::size_type motif_i = 0;
963 Sequence::size_type motif_len = a_motif.length();
964 Sequence::value_type motif_char;
965 Sequence::value_type seq_char;
967 while (seq_i < size())
969 // this is pretty much a straight translation of Nora's python code
970 // to match iupac letter codes
971 motif_char = toupper(a_motif[motif_i]);
972 seq_char = toupper(seq->at(seq_start+seq_i));
973 if (motif_char =='N')
975 else if (motif_char == seq_char)
977 else if ((motif_char =='M') && ((seq_char=='A') || (seq_char=='C')))
979 else if ((motif_char =='R') && ((seq_char=='A') || (seq_char=='G')))
981 else if ((motif_char =='W') && ((seq_char=='A') || (seq_char=='T')))
983 else if ((motif_char =='S') && ((seq_char=='C') || (seq_char=='G')))
985 else if ((motif_char =='Y') && ((seq_char=='C') || (seq_char=='T')))
987 else if ((motif_char =='K') && ((seq_char=='G') || (seq_char=='T')))
989 else if ((motif_char =='V') &&
990 ((seq_char=='A') || (seq_char=='C') || (seq_char=='G')))
992 else if ((motif_char =='H') &&
993 ((seq_char=='A') || (seq_char=='C') || (seq_char=='T')))
995 else if ((motif_char =='D') &&
996 ((seq_char=='A') || (seq_char=='G') || (seq_char=='T')))
998 else if ((motif_char =='B') &&
999 ((seq_char=='C') || (seq_char=='G') || (seq_char=='T')))
1003 // if a motif doesn't match, erase our current trial and try again
1008 // end Nora stuff, now we see if a match is found this pass
1009 if (motif_i == motif_len)
1012 motif_match_starts->push_back(seq_i - motif_len + 1);
1018 //std::cout << std::endl;
1021 void Sequence::add_string_annotation(std::string a_seq,
1024 std::vector<int> seq_starts = find_motif(a_seq);
1026 //std::cout << "searching for " << a_seq << " found " << seq_starts.size() << std::endl;
1028 for(std::vector<int>::iterator seq_start_i = seq_starts.begin();
1029 seq_start_i != seq_starts.end();
1032 annots.push_back(annot(*seq_start_i,
1033 *seq_start_i+a_seq.size(),
1039 void Sequence::find_sequences(std::list<Sequence>::iterator start,
1040 std::list<Sequence>::iterator end)
1042 while (start != end) {
1043 add_string_annotation(start->get_sequence(), start->get_fasta_header());
1049 std::ostream& operator<<(std::ostream& out, const Sequence& s)
1051 for(Sequence::const_iterator s_i = s.begin(); s_i != s.end(); ++s_i) {
1057 bool operator<(const Sequence& x, const Sequence& y)
1059 Sequence::const_iterator x_i = x.begin();
1060 Sequence::const_iterator y_i = y.begin();
1061 // for sequences there's some computation associated with computing .end
1062 // so lets cache it.
1063 Sequence::const_iterator xend = x.end();
1064 Sequence::const_iterator yend = y.end();
1066 if( x_i == xend and y_i == yend ) {
1068 } else if ( x_i == xend ) {
1070 } else if ( y_i == yend ) {
1072 } else if ( (*x_i) < (*y_i)) {
1074 } else if ( (*x_i) > (*y_i) ) {
1083 bool operator==(const Sequence& x, const Sequence& y)
1085 if (x.empty() and y.empty()) {
1086 // if there's no sequence in either sequence structure, they're equal
1088 } else if (x.empty() or y.empty()) {
1089 // if we fail the first test, and we discover one is empty,
1090 // we know they can't be equal. (and we need to do this
1091 // to prevent dereferencing an empty pointer)
1093 } else if (x.seq_count != y.seq_count) {
1094 // if they're of different lenghts, they're not equal
1097 Sequence::const_iterator xseq_i = x.begin();
1098 Sequence::const_iterator yseq_i = y.begin();
1099 // since the length of the two sequences is equal, we only need to
1101 for(; xseq_i != x.end(); ++xseq_i, ++yseq_i) {
1102 if (toupper(*xseq_i) != toupper(*yseq_i)) {
1109 bool operator!=(const Sequence& x, const Sequence& y)
1111 return not operator==(x, y);