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");
93 Sequence::Sequence(const char *seq)
97 set_filtered_sequence(seq);
100 Sequence::Sequence(const std::string& seq)
104 set_filtered_sequence(seq);
107 Sequence::Sequence(const Sequence& o)
112 motif_list(o.motif_list)
116 Sequence &Sequence::operator=(const Sequence& s)
119 //sequence = s.sequence;
124 motif_list = s.motif_list;
129 static void multiplatform_getline(std::istream& in, std::string& line)
134 while(in.good() and !(c == '\012' or c == '\015') ) {
138 // if we have cr-lf eat it
140 if (c=='\012' or c == '\015') {
145 //! load a fasta file into a sequence
147 * \param file_path the location of the fasta file in the filesystem
148 * \param seq_num which sequence in the file to load
149 * \param start_index starting position in the fasta sequence, 0 for beginning
150 * \param end_index ending position in the fasta sequence, 0 for end
151 * \return error message, empty string if no error. (gag!)
153 void Sequence::load_fasta(fs::path file_path, int seq_num,
154 int start_index, int end_index)
156 fs::fstream data_file;
157 data_file.open(file_path, std::ios::in);
159 if (!data_file.good())
161 throw mussa_load_error("Sequence File: "+file_path.string()+" not found");
164 load_fasta(data_file, seq_num, start_index, end_index);
165 } catch(sequence_empty_error e) {
166 // there doesn't appear to be any sequence
167 // catch and rethrow to include the filename
168 std::stringstream msg;
169 msg << "The selected sequence in "
170 << file_path.native_file_string()
171 << " appears to be empty";
172 throw sequence_empty_error(msg.str());
173 } catch(sequence_empty_file_error e) {
174 std::stringstream errormsg;
175 errormsg << file_path.native_file_string()
176 << " did not have any fasta sequences" << std::endl;
177 throw sequence_empty_file_error(errormsg.str());
183 Sequence::load_fasta(std::iostream& data_file, int seq_num,
184 int start_index, int end_index)
186 std::string file_data_line;
187 int header_counter = 0;
188 bool read_seq = true;
189 std::string rev_comp;
190 std::string sequence_raw;
191 std::string seq_tmp; // holds sequence during basic filtering
194 throw mussa_load_error("fasta sequence number is 1 based (can't be 0)");
197 // search for the header of the fasta sequence we want
198 while ( (!data_file.eof()) && (header_counter < seq_num) )
200 multiplatform_getline(data_file, file_data_line);
201 if (file_data_line.substr(0,1) == ">")
205 if (header_counter > 0) {
206 header = file_data_line.substr(1);
210 while ( !data_file.eof() && read_seq ) {
211 multiplatform_getline(data_file,file_data_line);
212 if (file_data_line.substr(0,1) == ">")
214 else sequence_raw += file_data_line;
217 // Lastly, if subselection of the sequence was specified we keep cut out
218 // and only keep that part
219 // end_index = 0 means no end was specified, so cut to the end
221 end_index = sequence_raw.size();
223 // sequence filtering for upcasing agctn and convert non AGCTN to N
224 if (end_index-start_index <= 0) {
225 std::string msg("The selected sequence appears to be empty");
226 throw sequence_empty_error(msg);
228 set_filtered_sequence(sequence_raw, start_index, end_index-start_index);
230 std::string errormsg("There were no fasta sequences");
231 throw sequence_empty_file_error(errormsg);
235 void Sequence::set_filtered_sequence(const std::string &old_seq,
236 std::string::size_type start,
237 std::string::size_type count)
239 char conversionTable[257];
242 count = old_seq.size() - start;
243 boost::shared_ptr<seq_string> new_seq(new seq_string);
244 new_seq->reserve(count);
246 // Make a conversion table
248 // everything we don't specify below will become 'N'
249 for(int table_i=0; table_i < 256; table_i++)
251 conversionTable[table_i] = 'N';
253 // add end of string character for printing out table for testing purposes
254 conversionTable[256] = '\0';
256 // we want these to map to themselves - ie not to change
257 conversionTable[(int)'A'] = 'A';
258 conversionTable[(int)'T'] = 'T';
259 conversionTable[(int)'G'] = 'G';
260 conversionTable[(int)'C'] = 'C';
262 conversionTable[(int)'a'] = 'A';
263 conversionTable[(int)'t'] = 'T';
264 conversionTable[(int)'g'] = 'G';
265 conversionTable[(int)'c'] = 'C';
267 // finally, the actual conversion loop
268 for(std::string::size_type seq_index = 0; seq_index < count; seq_index++)
270 new_seq->append(1, conversionTable[ (int)old_seq[seq_index+start]]);
276 Sequence::load_annot(fs::path file_path, int start_index, int end_index)
278 fs::fstream data_stream(file_path, std::ios::in);
281 throw mussa_load_error("Sequence File: " + file_path.string() + " not found");
284 // so i should probably be passing the parse function some iterators
285 // but the annotations files are (currently) small, so i think i can
286 // get away with loading the whole file into memory
289 while(data_stream.good()) {
295 parse_annot(data, start_index, end_index);
298 /* If this works, yikes, this is some brain hurting code.
300 * what's going on is that when pb_annot is instantiated it stores references
301 * to begin, end, name, type, declared in the parse function, then
302 * when operator() is called it grabs values from those references
303 * and uses that to instantiate an annot object and append that to our
306 * This weirdness is because the spirit library requires that actions
307 * conform to a specific prototype operator()(IteratorT, IteratorT)
308 * which doesn't provide any useful opportunity for me to actually
309 * grab the results of our parsing.
311 * so I instantiate this structure in order to have a place to grab
315 struct push_back_annot {
316 std::list<annot>& annot_list;
322 push_back_annot(std::list<annot>& annot_list_,
327 : annot_list(annot_list_),
335 void operator()(std::string::const_iterator,
336 std::string::const_iterator) const
338 //std::cout << "adding annot: " << begin << "|" << end << "|" << name << "|" << type << std::endl;
339 annot_list.push_back(annot(begin, end, name, type));
343 struct push_back_seq {
344 std::list<Sequence>& seq_list;
348 push_back_seq(std::list<Sequence>& seq_list_,
351 : seq_list(seq_list_),
357 void operator()(std::string::const_iterator,
358 std::string::const_iterator) const
360 // filter out newlines from our sequence
362 for(std::string::const_iterator seq_i = seq.begin();
366 if (*seq_i != '\015' && *seq_i != '\012') new_seq += *seq_i;
368 //std::cout << "adding seq: " << name << " " << new_seq << std::endl;
371 s.set_fasta_header(name);
372 seq_list.push_back(s);
377 Sequence::parse_annot(std::string data, int start_index, int end_index)
384 std::list<Sequence> query_seqs;
386 bool status = spirit::parse(data.begin(), data.end(),
393 )[spirit::assign_a(species)] >>
397 ( // ignore html tags
398 *(spirit::space_p) >>
400 +(~spirit::ch_p('>')) >>
405 ( // parse an absolute location name
406 (spirit::uint_p[spirit::assign_a(start)] >>
408 spirit::uint_p[spirit::assign_a(end)] >>
413 )[spirit::assign_a(name)] >>
420 )[spirit::assign_a(type)]
422 // to understand how this group gets set
423 // read the comment above struct push_back_annot
424 )[push_back_annot(annots, start, end, type, name)]
426 ((spirit::ch_p('>')|spirit::str_p(">")) >>
427 (*(spirit::print_p))[spirit::assign_a(name)] >>
429 (+(spirit::chset<>(nucleic_iupac_alphabet.c_str())))[spirit::assign_a(seq)]
430 )[push_back_seq(query_seqs, name, seq)]
436 spirit::space_p*/).full;
438 // go seearch for query sequences
439 find_sequences(query_seqs.begin(), query_seqs.end());
443 void Sequence::add_annotation(const annot& a)
448 const std::list<annot>& Sequence::annotations() const
454 Sequence::subseq(int start, int count) const
461 // there might be an off by one error with start+count > size()
462 if ( count == npos || start+count > size()) {
463 count = size()-start;
465 Sequence new_seq(seq->substr(start, count));
466 new_seq.set_fasta_header(get_fasta_header());
467 new_seq.set_species(get_species());
469 new_seq.motif_list = motif_list;
470 // attempt to copy & reannotate position based annotations
471 int end = start+count;
473 for(std::list<annot>::const_iterator annot_i = annots.begin();
474 annot_i != annots.end();
477 int annot_begin= annot_i->begin;
478 int annot_end = annot_i->end;
480 if (annot_begin < end) {
481 if (annot_begin >= start) {
482 annot_begin -= start;
487 if (annot_end < end) {
493 annot new_annot(annot_begin, annot_end, annot_i->type, annot_i->name);
494 new_seq.annots.push_back(new_annot);
502 Sequence::rev_comp() const
504 std::string rev_comp;
505 char conversionTable[257];
506 int seq_i, table_i, len;
509 rev_comp.reserve(len);
510 // make a conversion table
511 // init all parts of conversion table to '~' character
512 // '~' I doubt will ever appear in a sequence file (jeez, I hope)
513 // and may the fleas of 1000 camels infest the genitals of any biologist (and
514 // seven generations of their progeny) who decides to make it mean
515 // something special!!!
516 // PS - double the curse for any smartass non-biologist who tries it as well
517 for(table_i=0; table_i < 256; table_i++)
519 conversionTable[table_i] = '~';
521 // add end of string character for printing out table for testing purposes
522 conversionTable[256] = '\0';
524 // add in the characters for the bases we want to convert
525 conversionTable[(int)'A'] = 'T';
526 conversionTable[(int)'T'] = 'A';
527 conversionTable[(int)'G'] = 'C';
528 conversionTable[(int)'C'] = 'G';
529 conversionTable[(int)'N'] = 'N';
531 // finally, the actual conversion loop
532 for(seq_i = len - 1; seq_i >= 0; seq_i--)
534 table_i = (int) seq->at(seq_i);
535 rev_comp += conversionTable[table_i];
541 void Sequence::set_fasta_header(std::string header_)
546 void Sequence::set_species(const std::string& name)
551 std::string Sequence::get_species() const
558 Sequence::get_fasta_header() const
564 Sequence::get_name() const
566 if (header.size() > 0)
568 else if (species.size() > 0)
574 void Sequence::set_sequence(const std::string& s)
576 set_filtered_sequence(s);
579 std::string Sequence::get_sequence() const
584 Sequence::const_reference Sequence::operator[](Sequence::size_type i) const
589 Sequence::const_reference Sequence::at(Sequence::size_type i) const
591 if (!seq) throw std::out_of_range("empty sequence");
605 const char *Sequence::c_str() const
613 Sequence::const_iterator Sequence::begin() const
618 return Sequence::const_iterator(0);
621 Sequence::const_iterator Sequence::end() const
626 return Sequence::const_iterator(0);
629 bool Sequence::empty() const
637 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 // looks like the sequence is written as a single line
699 set_filtered_sequence(file_data_line);
700 getline(load_file, file_data_line);
701 getline(load_file, file_data_line);
702 if (file_data_line == "<Annotations>")
704 getline(load_file, file_data_line);
705 species = file_data_line;
706 while ( (!load_file.eof()) && (file_data_line != "</Annotations>") )
708 getline(load_file,file_data_line);
709 if ((file_data_line != "") && (file_data_line != "</Annotations>"))
711 // need to get 4 values...almost same code 4 times...
712 // get annot start index
713 space_split_i = file_data_line.find(" ");
714 annot_value = file_data_line.substr(0,space_split_i);
715 an_annot.begin = atoi (annot_value.c_str());
716 file_data_line = file_data_line.substr(space_split_i+1);
717 // get annot end index
718 space_split_i = file_data_line.find(" ");
719 annot_value = file_data_line.substr(0,space_split_i);
720 an_annot.end = atoi (annot_value.c_str());
722 if (space_split_i == std::string::npos) // no entry for type or name
724 std::cout << "seq, annots - no type or name\n";
728 else // else get annot type
730 file_data_line = file_data_line.substr(space_split_i+1);
731 space_split_i = file_data_line.find(" ");
732 annot_value = file_data_line.substr(0,space_split_i);
733 an_annot.type = annot_value;
734 if (space_split_i == std::string::npos) // no entry for name
736 std::cout << "seq, annots - no name\n";
739 else // get annot name
741 file_data_line = file_data_line.substr(space_split_i+1);
742 space_split_i = file_data_line.find(" ");
743 annot_value = file_data_line.substr(0,space_split_i);
744 an_annot.type = annot_value;
747 annots.push_back(an_annot); // don't forget to actually add the annot
749 //std::cout << "seq, annots: " << an_annot.start << ", " << an_annot.end
750 // << "-->" << an_annot.type << "::" << an_annot.name << std::endl;
757 Sequence::rc_motif(std::string a_motif) const
759 std::string rev_comp;
760 char conversionTable[257];
761 int seq_i, table_i, len;
763 len = a_motif.length();
764 rev_comp.reserve(len);
766 for(table_i=0; table_i < 256; table_i++)
768 conversionTable[table_i] = '~';
770 // add end of std::string character for printing out table for testing purposes
771 conversionTable[256] = '\0';
773 // add in the characters for the bases we want to convert (IUPAC)
774 conversionTable[(int)'A'] = 'T';
775 conversionTable[(int)'T'] = 'A';
776 conversionTable[(int)'G'] = 'C';
777 conversionTable[(int)'C'] = 'G';
778 conversionTable[(int)'N'] = 'N';
779 conversionTable[(int)'M'] = 'K';
780 conversionTable[(int)'R'] = 'Y';
781 conversionTable[(int)'W'] = 'W';
782 conversionTable[(int)'S'] = 'S';
783 conversionTable[(int)'Y'] = 'R';
784 conversionTable[(int)'K'] = 'M';
785 conversionTable[(int)'V'] = 'B';
786 conversionTable[(int)'H'] = 'D';
787 conversionTable[(int)'D'] = 'H';
788 conversionTable[(int)'B'] = 'V';
790 // finally, the actual conversion loop
791 for(seq_i = len - 1; seq_i >= 0; seq_i--)
793 //std::cout << "** i = " << seq_i << " bp = " <<
794 table_i = (int) a_motif[seq_i];
795 rev_comp += conversionTable[table_i];
798 //std::cout << "seq: " << a_motif << std::endl;
799 //std::cout << "rc: " << rev_comp << std::endl;
805 Sequence::motif_normalize(const std::string& a_motif)
807 std::string valid_motif;
810 len = a_motif.length();
811 valid_motif.reserve(len);
813 // this just upcases IUPAC symbols. Eventually should return an error if non IUPAC is present.
814 // current nonIUPAC symbols are omitted, which is not reported atm
815 for(seq_i = 0; seq_i < len; seq_i++)
817 if ((a_motif[seq_i] == 'a') || (a_motif[seq_i] == 'A'))
819 else if ((a_motif[seq_i] == 't') || (a_motif[seq_i] == 'T'))
821 else if ((a_motif[seq_i] == 'g') || (a_motif[seq_i] == 'G'))
823 else if ((a_motif[seq_i] == 'c') || (a_motif[seq_i] == 'C'))
825 else if ((a_motif[seq_i] == 'n') || (a_motif[seq_i] == 'N'))
827 else if ((a_motif[seq_i] == 'm') || (a_motif[seq_i] == 'M'))
829 else if ((a_motif[seq_i] == 'r') || (a_motif[seq_i] == 'R'))
831 else if ((a_motif[seq_i] == 'w') || (a_motif[seq_i] == 'W'))
833 else if ((a_motif[seq_i] == 's') || (a_motif[seq_i] == 'S'))
835 else if ((a_motif[seq_i] == 'y') || (a_motif[seq_i] == 'Y'))
837 else if ((a_motif[seq_i] == 'k') || (a_motif[seq_i] == 'K'))
839 else if ((a_motif[seq_i] == 'v') || (a_motif[seq_i] == 'V'))
841 else if ((a_motif[seq_i] == 'h') || (a_motif[seq_i] == 'H'))
843 else if ((a_motif[seq_i] == 'd') || (a_motif[seq_i] == 'D'))
845 else if ((a_motif[seq_i] == 'b') || (a_motif[seq_i] == 'B'))
848 std::string msg = "Letter ";
849 msg += a_motif[seq_i];
850 msg += " is not a valid IUPAC symbol";
851 throw motif_normalize_error(msg);
854 //std::cout << "valid_motif is: " << valid_motif << std::endl;
858 void Sequence::add_motif(const Sequence& a_motif)
860 std::vector<int> motif_starts = find_motif(a_motif);
862 for(std::vector<int>::iterator motif_start_i = motif_starts.begin();
863 motif_start_i != motif_starts.end();
866 motif_list.push_back(motif(*motif_start_i, a_motif.get_sequence()));
870 void Sequence::clear_motifs()
875 const std::list<motif>& Sequence::motifs() const
881 Sequence::find_motif(const std::string& a_motif) const
883 std::vector<int> motif_match_starts;
884 std::string norm_motif_rc;
886 motif_match_starts.clear();
888 //std::cout << "motif is: " << a_motif << std::endl;
889 std::string norm_motif = motif_normalize(a_motif);
890 //std::cout << "motif is: " << a_motif << std::endl;
892 if (norm_motif.size() > 0)
894 //std::cout << "Sequence: none blank motif\n";
895 motif_scan(norm_motif, &motif_match_starts);
897 norm_motif_rc = rc_motif(a_motif);
898 // make sure not to do search again if it is a palindrome
899 if (norm_motif_rc != norm_motif) {
900 motif_scan(norm_motif_rc, &motif_match_starts);
903 return motif_match_starts;
907 Sequence::find_motif(const Sequence& a_motif) const
909 return find_motif(a_motif.get_sequence());
913 Sequence::motif_scan(std::string a_motif, std::vector<int> * motif_match_starts) const
915 // if there's no sequence we can't scan for it?
916 // should this throw an exception?
919 std::string::const_iterator seq_c = seq->begin();
920 std::string::size_type seq_i;
921 int motif_i, motif_len;
923 //std::cout << "Sequence: motif, seq len = " << sequence.length() << std::endl;
924 motif_len = a_motif.length();
926 //std::cout << "motif_length: " << motif_len << std::endl;
927 //std::cout << "RAAARRRRR\n";
931 //std::cout << "motif: " << a_motif << std::endl;
933 //std::cout << "Sequence: motif, length= " << length << std::endl;
935 while (seq_i < length())
937 //std::cout << seq_c[seq_i];
938 //std::cout << seq_c[seq_i] << "?" << a_motif[motif_i] << ":" << motif_i << " ";
939 // this is pretty much a straight translation of Nora's python code
940 // to match iupac letter codes
941 if (a_motif[motif_i] =='N')
943 else if (a_motif[motif_i] == seq_c[seq_i])
945 else if ((a_motif[motif_i] =='M') &&
946 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C')))
948 else if ((a_motif[motif_i] =='R') &&
949 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G')))
951 else if ((a_motif[motif_i] =='W') &&
952 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='T')))
954 else if ((a_motif[motif_i] =='S') &&
955 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G')))
957 else if ((a_motif[motif_i] =='Y') &&
958 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='T')))
960 else if ((a_motif[motif_i] =='K') &&
961 ((seq_c[seq_i]=='G') || (seq_c[seq_i]=='T')))
963 else if ((a_motif[motif_i] =='V') &&
964 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
965 (seq_c[seq_i]=='G')))
967 else if ((a_motif[seq_i] =='H') &&
968 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
969 (seq_c[seq_i]=='T')))
971 else if ((a_motif[motif_i] =='D') &&
972 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G') ||
973 (seq_c[seq_i]=='T')))
975 else if ((a_motif[motif_i] =='B') &&
976 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G') ||
977 (seq_c[seq_i]=='T')))
986 // end Nora stuff, now we see if a match is found this pass
987 if (motif_i == motif_len)
991 motif_match_starts->push_back(seq_i - motif_len + 1);
997 //std::cout << std::endl;
1000 void Sequence::add_string_annotation(std::string a_seq,
1003 std::vector<int> seq_starts = find_motif(a_seq);
1005 //std::cout << "searching for " << a_seq << " found " << seq_starts.size() << std::endl;
1007 for(std::vector<int>::iterator seq_start_i = seq_starts.begin();
1008 seq_start_i != seq_starts.end();
1011 annots.push_back(annot(*seq_start_i,
1012 *seq_start_i+a_seq.size(),
1018 void Sequence::find_sequences(std::list<Sequence>::iterator start,
1019 std::list<Sequence>::iterator end)
1021 while (start != end) {
1022 add_string_annotation(start->get_sequence(), start->get_fasta_header());
1028 std::ostream& operator<<(std::ostream& out, const Sequence& s)
1030 for(Sequence::const_iterator s_i = s.begin(); s_i != s.end(); ++s_i) {
1036 bool operator<(const Sequence& x, const Sequence& y)
1038 Sequence::const_iterator x_i = x.begin();
1039 Sequence::const_iterator y_i = y.begin();
1042 if( x_i == x.end() and y_i == y.end() ) {
1044 } else if ( x_i == x.end() ) {
1046 } else if ( y_i == y.end() ) {
1048 } else if ( (*x_i) < (*y_i)) {
1050 } else if ( (*x_i) > (*y_i) ) {
1059 bool operator==(const Sequence& x, const Sequence& y)
1061 if (x.empty() and y.empty()) {
1062 // if there's no sequence in either sequence structure, they're equal
1064 } else if (x.empty() or y.empty()) {
1065 // if we fail the first test, and we discover one is empty,
1066 // we know they can't be equal. (and we need to do this
1067 // to prevent dereferencing an empty pointer)
1069 } else if ( *(x.seq) == *(y.seq)) {
1070 // and x.annots == y.annots and x.motif_list == y.motif_list) {