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"
48 annot::annot(int begin, int end, std::string type, std::string name)
60 bool operator==(const annot& left, const annot& right)
62 return ((left.begin== right.begin) and
63 (left.end == right.end) and
64 (left.type == right.type) and
65 (left.name == right.name));
68 motif::motif(int begin, std::string motif)
69 : annot(begin, begin+motif.size(), "motif", motif),
78 const std::string Sequence::dna_alphabet("AaCcGgTtNn\012\015");
79 const std::string Sequence::rna_alphabet("AaCcGgNnUu\012\015");
80 //! this is the general iupac alphabet for nucleotides
81 const std::string Sequence::nucleic_iupac_alphabet("AaCcGgTtUuRrYyMmKkSsWwBbDdHhVvNn\012\015");
82 //! the protein alphabet
83 const std::string Sequence::protein_alphabet("AaCcDdEeFfGgHhIiKkLlMmNnPpQqRrSsTtVvWwYy\012\015");
96 Sequence::Sequence(const char *seq)
100 set_filtered_sequence(seq);
103 Sequence::Sequence(const std::string& seq)
107 set_filtered_sequence(seq);
110 Sequence::Sequence(const Sequence& o)
115 motif_list(o.motif_list)
119 Sequence &Sequence::operator=(const Sequence& s)
122 //sequence = s.sequence;
127 motif_list = s.motif_list;
133 Sequence::operator std::string()
135 std::string s(seq.begin(), seq.end());
139 Sequence::operator std::string() const
141 std::string s(seq.begin(), seq.end());
146 static void multiplatform_getline(std::istream& in, std::string& line)
151 while(in.good() and !(c == '\012' or c == '\015') ) {
155 // if we have cr-lf eat it
157 if (c=='\012' or c == '\015') {
162 //! load a fasta file into a sequence
164 * \param file_path the location of the fasta file in the filesystem
165 * \param seq_num which sequence in the file to load
166 * \param start_index starting position in the fasta sequence, 0 for beginning
167 * \param end_index ending position in the fasta sequence, 0 for end
168 * \return error message, empty string if no error. (gag!)
170 void Sequence::load_fasta(fs::path file_path, int seq_num,
171 int start_index, int end_index)
173 fs::fstream data_file;
174 data_file.open(file_path, std::ios::in);
176 if (!data_file.good())
178 throw mussa_load_error("Sequence File: "+file_path.string()+" not found");
181 load_fasta(data_file, seq_num, start_index, end_index);
182 } catch(sequence_empty_error e) {
183 // there doesn't appear to be any sequence
184 // catch and rethrow to include the filename
185 std::stringstream msg;
186 msg << "The selected sequence in "
187 << file_path.native_file_string()
188 << " appears to be empty";
189 throw sequence_empty_error(msg.str());
190 } catch(sequence_empty_file_error e) {
191 std::stringstream errormsg;
192 errormsg << file_path.native_file_string()
193 << " did not have any fasta sequences" << std::endl;
194 throw sequence_empty_file_error(errormsg.str());
200 Sequence::load_fasta(std::iostream& data_file, int seq_num,
201 int start_index, int end_index)
203 std::string file_data_line;
204 int header_counter = 0;
205 bool read_seq = true;
206 std::string rev_comp;
207 std::string sequence_raw;
208 std::string seq_tmp; // holds sequence during basic filtering
211 throw mussa_load_error("fasta sequence number is 1 based (can't be 0)");
214 // search for the header of the fasta sequence we want
215 while ( (!data_file.eof()) && (header_counter < seq_num) )
217 multiplatform_getline(data_file, file_data_line);
218 if (file_data_line.substr(0,1) == ">")
222 if (header_counter > 0) {
223 header = file_data_line.substr(1);
227 while ( !data_file.eof() && read_seq ) {
228 multiplatform_getline(data_file,file_data_line);
229 if (file_data_line.substr(0,1) == ">")
231 else sequence_raw += file_data_line;
234 // Lastly, if subselection of the sequence was specified we keep cut out
235 // and only keep that part
236 // end_index = 0 means no end was specified, so cut to the end
238 end_index = sequence_raw.size();
240 // sequence filtering for upcasing agctn and convert non AGCTN to N
241 if (end_index-start_index <= 0) {
242 std::string msg("The selected sequence appears to be empty");
243 throw sequence_empty_error(msg);
245 set_filtered_sequence(sequence_raw, start_index, end_index-start_index);
247 std::string errormsg("There were no fasta sequences");
248 throw sequence_empty_file_error(errormsg);
252 void Sequence::set_filtered_sequence(const std::string &old_seq,
253 std::string::size_type start,
254 std::string::size_type count)
256 char conversionTable[257];
259 count = old_seq.size() - start;
263 // Make a conversion table
265 // everything we don't specify below will become 'N'
266 for(int table_i=0; table_i < 256; table_i++)
268 conversionTable[table_i] = 'N';
270 // add end of string character for printing out table for testing purposes
271 conversionTable[256] = '\0';
273 // we want these to map to themselves - ie not to change
274 conversionTable[(int)'A'] = 'A';
275 conversionTable[(int)'T'] = 'T';
276 conversionTable[(int)'G'] = 'G';
277 conversionTable[(int)'C'] = 'C';
279 conversionTable[(int)'a'] = 'A';
280 conversionTable[(int)'t'] = 'T';
281 conversionTable[(int)'g'] = 'G';
282 conversionTable[(int)'c'] = 'C';
284 // finally, the actual conversion loop
285 for(std::string::size_type seq_index = 0; seq_index < count; seq_index++)
287 seq.append(1, conversionTable[ (int)old_seq[seq_index+start]]);
292 Sequence::load_annot(fs::path file_path, int start_index, int end_index)
294 fs::fstream data_stream(file_path, std::ios::in);
297 throw mussa_load_error("Sequence File: " + file_path.string() + " not found");
300 // so i should probably be passing the parse function some iterators
301 // but the annotations files are (currently) small, so i think i can
302 // get away with loading the whole file into memory
305 while(data_stream.good()) {
311 parse_annot(data, start_index, end_index);
314 /* If this works, yikes, this is some brain hurting code.
316 * what's going on is that when pb_annot is instantiated it stores references
317 * to begin, end, name, type, declared in the parse function, then
318 * when operator() is called it grabs values from those references
319 * and uses that to instantiate an annot object and append that to our
322 * This weirdness is because the spirit library requires that actions
323 * conform to a specific prototype operator()(IteratorT, IteratorT)
324 * which doesn't provide any useful opportunity for me to actually
325 * grab the results of our parsing.
327 * so I instantiate this structure in order to have a place to grab
331 struct push_back_annot {
332 std::list<annot>& annot_list;
338 push_back_annot(std::list<annot>& annot_list_,
343 : annot_list(annot_list_),
351 void operator()(std::string::const_iterator,
352 std::string::const_iterator) const
354 //std::cout << "adding annot: " << begin << "|" << end << "|" << name << "|" << type << std::endl;
355 annot_list.push_back(annot(begin, end, name, type));
359 struct push_back_seq {
360 std::list<Sequence>& seq_list;
364 push_back_seq(std::list<Sequence>& seq_list_,
367 : seq_list(seq_list_),
373 void operator()(std::string::const_iterator,
374 std::string::const_iterator) const
376 // filter out newlines from our sequence
378 for(std::string::const_iterator seq_i = seq.begin();
382 if (*seq_i != '\015' && *seq_i != '\012') new_seq += *seq_i;
384 //std::cout << "adding seq: " << name << " " << new_seq << std::endl;
387 s.set_fasta_header(name);
388 seq_list.push_back(s);
393 Sequence::parse_annot(std::string data, int start_index, int end_index)
400 std::list<Sequence> query_seqs;
402 bool status = spirit::parse(data.begin(), data.end(),
409 )[spirit::assign_a(species)] >>
413 ( // ignore html tags
414 *(spirit::space_p) >>
416 +(~spirit::ch_p('>')) >>
421 ( // parse an absolute location name
422 (spirit::uint_p[spirit::assign_a(start)] >>
424 spirit::uint_p[spirit::assign_a(end)] >>
429 )[spirit::assign_a(name)] >>
436 )[spirit::assign_a(type)]
438 // to understand how this group gets set
439 // read the comment above struct push_back_annot
440 )[push_back_annot(annots, start, end, type, name)]
442 ((spirit::ch_p('>')|spirit::str_p(">")) >>
443 (*(spirit::print_p))[spirit::assign_a(name)] >>
445 (+(spirit::chset<>(nucleic_iupac_alphabet.c_str())))[spirit::assign_a(seq)]
446 )[push_back_seq(query_seqs, name, seq)]
452 spirit::space_p*/).full;
454 // go seearch for query sequences
455 find_sequences(query_seqs.begin(), query_seqs.end());
459 void Sequence::add_annotation(const annot& a)
464 const std::list<annot>& Sequence::annotations() const
470 Sequence::subseq(int start, int count) const
472 // there might be an off by one error with start+count > size()
473 if ( count == npos || start+count > size()) {
474 count = size()-start;
476 Sequence new_seq(seq.substr(start, count));
477 new_seq.set_fasta_header(get_fasta_header());
478 new_seq.set_species(get_species());
480 new_seq.motif_list = motif_list;
481 // attempt to copy & reannotate position based annotations
482 int end = start+count;
484 for(std::list<annot>::const_iterator annot_i = annots.begin();
485 annot_i != annots.end();
488 int annot_begin= annot_i->begin;
489 int annot_end = annot_i->end;
491 if (annot_begin < end) {
492 if (annot_begin >= start) {
493 annot_begin -= start;
498 if (annot_end < end) {
504 annot new_annot(annot_begin, annot_end, annot_i->type, annot_i->name);
505 new_seq.annots.push_back(new_annot);
513 Sequence::rev_comp() const
515 std::string rev_comp;
516 char conversionTable[257];
517 int seq_i, table_i, len;
520 rev_comp.reserve(len);
521 // make a conversion table
522 // init all parts of conversion table to '~' character
523 // '~' I doubt will ever appear in a sequence file (jeez, I hope)
524 // and may the fleas of 1000 camels infest the genitals of any biologist (and
525 // seven generations of their progeny) who decides to make it mean
526 // something special!!!
527 // PS - double the curse for any smartass non-biologist who tries it as well
528 for(table_i=0; table_i < 256; table_i++)
530 conversionTable[table_i] = '~';
532 // add end of string character for printing out table for testing purposes
533 conversionTable[256] = '\0';
535 // add in the characters for the bases we want to convert
536 conversionTable[(int)'A'] = 'T';
537 conversionTable[(int)'T'] = 'A';
538 conversionTable[(int)'G'] = 'C';
539 conversionTable[(int)'C'] = 'G';
540 conversionTable[(int)'N'] = 'N';
542 // finally, the actual conversion loop
543 for(seq_i = len - 1; seq_i >= 0; seq_i--)
545 table_i = (int) seq.at(seq_i);
546 rev_comp += conversionTable[table_i];
552 void Sequence::set_fasta_header(std::string header_)
557 void Sequence::set_species(const std::string& name)
562 std::string Sequence::get_species() const
569 Sequence::get_fasta_header() const
575 Sequence::get_name() const
577 if (header.size() > 0)
579 else if (species.size() > 0)
585 void Sequence::set_sequence(const std::string& s)
587 set_filtered_sequence(s);
590 std::string Sequence::get_sequence() const
595 Sequence::const_reference Sequence::operator[](Sequence::size_type i) const
600 Sequence::const_reference Sequence::at(Sequence::size_type n) const
615 Sequence::iterator Sequence::begin()
620 Sequence::const_iterator Sequence::begin() const
625 Sequence::iterator Sequence::end()
630 Sequence::const_iterator Sequence::end() const
635 bool Sequence::empty() const
640 Sequence::size_type Sequence::size() const
645 Sequence::size_type Sequence::length() const
650 Sequence::save(fs::fstream &save_file)
651 //std::string save_file_path)
654 std::list<annot>::iterator annots_i;
656 // not sure why, or if i'm doing something wrong, but can't seem to pass
657 // file pointers down to this method from the mussa control class
658 // so each call to save a sequence appends to the file started by mussa_class
659 //save_file.open(save_file_path.c_str(), std::ios::app);
661 save_file << "<Sequence>" << std::endl;
662 save_file << *this << std::endl;
663 save_file << "</Sequence>" << std::endl;
665 save_file << "<Annotations>" << std::endl;
666 save_file << species << std::endl;
667 for (annots_i = annots.begin(); annots_i != annots.end(); ++annots_i)
669 save_file << annots_i->begin << " " << annots_i->end << " " ;
670 save_file << annots_i->name << " " << annots_i->type << std::endl;
672 save_file << "</Annotations>" << std::endl;
677 Sequence::load_museq(fs::path load_file_path, int seq_num)
679 fs::fstream load_file;
680 std::string file_data_line;
683 std::string::size_type space_split_i;
684 std::string annot_value;
687 load_file.open(load_file_path, std::ios::in);
690 // search for the seq_num-th sequence
691 while ( (!load_file.eof()) && (seq_counter < seq_num) )
693 getline(load_file,file_data_line);
694 if (file_data_line == "<Sequence>")
697 getline(load_file, file_data_line);
698 seq.assign(file_data_line);
699 getline(load_file, file_data_line);
700 getline(load_file, file_data_line);
701 if (file_data_line == "<Annotations>")
703 getline(load_file, file_data_line);
704 species = file_data_line;
705 while ( (!load_file.eof()) && (file_data_line != "</Annotations>") )
707 getline(load_file,file_data_line);
708 if ((file_data_line != "") && (file_data_line != "</Annotations>"))
710 // need to get 4 values...almost same code 4 times...
711 // get annot start index
712 space_split_i = file_data_line.find(" ");
713 annot_value = file_data_line.substr(0,space_split_i);
714 an_annot.begin = atoi (annot_value.c_str());
715 file_data_line = file_data_line.substr(space_split_i+1);
716 // get annot end index
717 space_split_i = file_data_line.find(" ");
718 annot_value = file_data_line.substr(0,space_split_i);
719 an_annot.end = atoi (annot_value.c_str());
721 if (space_split_i == std::string::npos) // no entry for type or name
723 std::cout << "seq, annots - no type or name\n";
727 else // else get annot type
729 file_data_line = file_data_line.substr(space_split_i+1);
730 space_split_i = file_data_line.find(" ");
731 annot_value = file_data_line.substr(0,space_split_i);
732 an_annot.type = annot_value;
733 if (space_split_i == std::string::npos) // no entry for name
735 std::cout << "seq, annots - no name\n";
738 else // get annot name
740 file_data_line = file_data_line.substr(space_split_i+1);
741 space_split_i = file_data_line.find(" ");
742 annot_value = file_data_line.substr(0,space_split_i);
743 an_annot.type = annot_value;
746 annots.push_back(an_annot); // don't forget to actually add the annot
748 //std::cout << "seq, annots: " << an_annot.start << ", " << an_annot.end
749 // << "-->" << an_annot.type << "::" << an_annot.name << std::endl;
756 Sequence::rc_motif(std::string a_motif) const
758 std::string rev_comp;
759 char conversionTable[257];
760 int seq_i, table_i, len;
762 len = a_motif.length();
763 rev_comp.reserve(len);
765 for(table_i=0; table_i < 256; table_i++)
767 conversionTable[table_i] = '~';
769 // add end of std::string character for printing out table for testing purposes
770 conversionTable[256] = '\0';
772 // add in the characters for the bases we want to convert (IUPAC)
773 conversionTable[(int)'A'] = 'T';
774 conversionTable[(int)'T'] = 'A';
775 conversionTable[(int)'G'] = 'C';
776 conversionTable[(int)'C'] = 'G';
777 conversionTable[(int)'N'] = 'N';
778 conversionTable[(int)'M'] = 'K';
779 conversionTable[(int)'R'] = 'Y';
780 conversionTable[(int)'W'] = 'W';
781 conversionTable[(int)'S'] = 'S';
782 conversionTable[(int)'Y'] = 'R';
783 conversionTable[(int)'K'] = 'M';
784 conversionTable[(int)'V'] = 'B';
785 conversionTable[(int)'H'] = 'D';
786 conversionTable[(int)'D'] = 'H';
787 conversionTable[(int)'B'] = 'V';
789 // finally, the actual conversion loop
790 for(seq_i = len - 1; seq_i >= 0; seq_i--)
792 //std::cout << "** i = " << seq_i << " bp = " <<
793 table_i = (int) a_motif[seq_i];
794 rev_comp += conversionTable[table_i];
797 //std::cout << "seq: " << a_motif << std::endl;
798 //std::cout << "rc: " << rev_comp << std::endl;
804 Sequence::motif_normalize(const std::string& a_motif)
806 std::string valid_motif;
809 len = a_motif.length();
810 valid_motif.reserve(len);
812 // this just upcases IUPAC symbols. Eventually should return an error if non IUPAC is present.
813 // current nonIUPAC symbols are omitted, which is not reported atm
814 for(seq_i = 0; seq_i < len; seq_i++)
816 if ((a_motif[seq_i] == 'a') || (a_motif[seq_i] == 'A'))
818 else if ((a_motif[seq_i] == 't') || (a_motif[seq_i] == 'T'))
820 else if ((a_motif[seq_i] == 'g') || (a_motif[seq_i] == 'G'))
822 else if ((a_motif[seq_i] == 'c') || (a_motif[seq_i] == 'C'))
824 else if ((a_motif[seq_i] == 'n') || (a_motif[seq_i] == 'N'))
826 else if ((a_motif[seq_i] == 'm') || (a_motif[seq_i] == 'M'))
828 else if ((a_motif[seq_i] == 'r') || (a_motif[seq_i] == 'R'))
830 else if ((a_motif[seq_i] == 'w') || (a_motif[seq_i] == 'W'))
832 else if ((a_motif[seq_i] == 's') || (a_motif[seq_i] == 'S'))
834 else if ((a_motif[seq_i] == 'y') || (a_motif[seq_i] == 'Y'))
836 else if ((a_motif[seq_i] == 'k') || (a_motif[seq_i] == 'K'))
838 else if ((a_motif[seq_i] == 'v') || (a_motif[seq_i] == 'V'))
840 else if ((a_motif[seq_i] == 'h') || (a_motif[seq_i] == 'H'))
842 else if ((a_motif[seq_i] == 'd') || (a_motif[seq_i] == 'D'))
844 else if ((a_motif[seq_i] == 'b') || (a_motif[seq_i] == 'B'))
847 std::string msg = "Letter ";
848 msg += a_motif[seq_i];
849 msg += " is not a valid IUPAC symbol";
850 throw motif_normalize_error(msg);
853 //std::cout << "valid_motif is: " << valid_motif << std::endl;
857 void Sequence::add_motif(const Sequence& a_motif)
859 std::vector<int> motif_starts = find_motif(a_motif);
861 for(std::vector<int>::iterator motif_start_i = motif_starts.begin();
862 motif_start_i != motif_starts.end();
865 motif_list.push_back(motif(*motif_start_i, a_motif.get_sequence()));
869 void Sequence::clear_motifs()
874 const std::list<motif>& Sequence::motifs() const
880 Sequence::find_motif(const std::string& a_motif) const
882 std::vector<int> motif_match_starts;
883 std::string norm_motif_rc;
885 motif_match_starts.clear();
887 //std::cout << "motif is: " << a_motif << std::endl;
888 std::string norm_motif = motif_normalize(a_motif);
889 //std::cout << "motif is: " << a_motif << std::endl;
891 if (norm_motif.size() > 0)
893 //std::cout << "Sequence: none blank motif\n";
894 motif_scan(norm_motif, &motif_match_starts);
896 norm_motif_rc = rc_motif(a_motif);
897 // make sure not to do search again if it is a palindrome
898 if (norm_motif_rc != norm_motif) {
899 motif_scan(norm_motif_rc, &motif_match_starts);
902 return motif_match_starts;
906 Sequence::find_motif(const Sequence& a_motif) const
908 return find_motif(a_motif.get_sequence());
912 Sequence::motif_scan(std::string a_motif, std::vector<int> * motif_match_starts) const
914 std::string::const_iterator seq_c = seq.begin();
915 std::string::size_type seq_i;
916 int motif_i, motif_len;
918 //std::cout << "Sequence: motif, seq len = " << sequence.length() << std::endl;
919 motif_len = a_motif.length();
921 //std::cout << "motif_length: " << motif_len << std::endl;
922 //std::cout << "RAAARRRRR\n";
926 //std::cout << "motif: " << a_motif << std::endl;
928 //std::cout << "Sequence: motif, length= " << length << std::endl;
930 while (seq_i < length())
932 //std::cout << seq_c[seq_i];
933 //std::cout << seq_c[seq_i] << "?" << a_motif[motif_i] << ":" << motif_i << " ";
934 // this is pretty much a straight translation of Nora's python code
935 // to match iupac letter codes
936 if (a_motif[motif_i] =='N')
938 else if (a_motif[motif_i] == seq_c[seq_i])
940 else if ((a_motif[motif_i] =='M') &&
941 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C')))
943 else if ((a_motif[motif_i] =='R') &&
944 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G')))
946 else if ((a_motif[motif_i] =='W') &&
947 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='T')))
949 else if ((a_motif[motif_i] =='S') &&
950 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G')))
952 else if ((a_motif[motif_i] =='Y') &&
953 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='T')))
955 else if ((a_motif[motif_i] =='K') &&
956 ((seq_c[seq_i]=='G') || (seq_c[seq_i]=='T')))
958 else if ((a_motif[motif_i] =='V') &&
959 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
960 (seq_c[seq_i]=='G')))
962 else if ((a_motif[seq_i] =='H') &&
963 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
964 (seq_c[seq_i]=='T')))
966 else if ((a_motif[motif_i] =='D') &&
967 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G') ||
968 (seq_c[seq_i]=='T')))
970 else if ((a_motif[motif_i] =='B') &&
971 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G') ||
972 (seq_c[seq_i]=='T')))
981 // end Nora stuff, now we see if a match is found this pass
982 if (motif_i == motif_len)
986 motif_match_starts->push_back(seq_i - motif_len + 1);
992 //std::cout << std::endl;
995 void Sequence::add_string_annotation(std::string a_seq,
998 std::vector<int> seq_starts = find_motif(a_seq);
1000 //std::cout << "searching for " << a_seq << " found " << seq_starts.size() << std::endl;
1002 for(std::vector<int>::iterator seq_start_i = seq_starts.begin();
1003 seq_start_i != seq_starts.end();
1006 annots.push_back(annot(*seq_start_i,
1007 *seq_start_i+a_seq.size(),
1013 void Sequence::find_sequences(std::list<Sequence>::iterator start,
1014 std::list<Sequence>::iterator end)
1016 while (start != end) {
1017 add_string_annotation(start->get_sequence(), start->get_fasta_header());
1023 std::ostream& operator<<(std::ostream& out, const Sequence& s)
1025 for(Sequence::const_iterator s_i = s.begin(); s_i != s.end(); ++s_i) {
1031 bool operator<(const Sequence& x, const Sequence& y)
1033 Sequence::const_iterator x_i = x.begin();
1034 Sequence::const_iterator y_i = y.begin();
1037 if( x_i == x.end() and y_i == y.end() ) {
1039 } else if ( x_i == x.end() ) {
1041 } else if ( y_i == y.end() ) {
1043 } else if ( (*x_i) < (*y_i)) {
1045 } else if ( (*x_i) > (*y_i) ) {
1054 bool operator==(const Sequence& x, const Sequence& y)
1056 if (x.seq == y.seq and x.annots == y.annots and x.motif_list == y.motif_list) {