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;
130 Sequence::operator std::string()
132 std::string s(seq.begin(), seq.end());
136 Sequence::operator std::string() const
138 std::string s(seq.begin(), seq.end());
143 static void multiplatform_getline(std::istream& in, std::string& line)
148 while(in.good() and !(c == '\012' or c == '\015') ) {
152 // if we have cr-lf eat it
154 if (c=='\012' or c == '\015') {
159 //! load a fasta file into a sequence
161 * \param file_path the location of the fasta file in the filesystem
162 * \param seq_num which sequence in the file to load
163 * \param start_index starting position in the fasta sequence, 0 for beginning
164 * \param end_index ending position in the fasta sequence, 0 for end
165 * \return error message, empty string if no error. (gag!)
167 void Sequence::load_fasta(fs::path file_path, int seq_num,
168 int start_index, int end_index)
170 fs::fstream data_file;
171 data_file.open(file_path, std::ios::in);
173 if (!data_file.good())
175 throw mussa_load_error("Sequence File: "+file_path.string()+" not found");
178 load_fasta(data_file, seq_num, start_index, end_index);
179 } catch(sequence_empty_error e) {
180 // there doesn't appear to be any sequence
181 // catch and rethrow to include the filename
182 std::stringstream msg;
183 msg << "The selected sequence in "
184 << file_path.native_file_string()
185 << " appears to be empty";
186 throw sequence_empty_error(msg.str());
187 } catch(sequence_empty_file_error e) {
188 std::stringstream errormsg;
189 errormsg << file_path.native_file_string()
190 << " did not have any fasta sequences" << std::endl;
191 throw sequence_empty_file_error(errormsg.str());
197 Sequence::load_fasta(std::iostream& data_file, int seq_num,
198 int start_index, int end_index)
200 std::string file_data_line;
201 int header_counter = 0;
202 bool read_seq = true;
203 std::string rev_comp;
204 std::string sequence_raw;
205 std::string seq_tmp; // holds sequence during basic filtering
208 throw mussa_load_error("fasta sequence number is 1 based (can't be 0)");
211 // search for the header of the fasta sequence we want
212 while ( (!data_file.eof()) && (header_counter < seq_num) )
214 multiplatform_getline(data_file, file_data_line);
215 if (file_data_line.substr(0,1) == ">")
219 if (header_counter > 0) {
220 header = file_data_line.substr(1);
224 while ( !data_file.eof() && read_seq ) {
225 multiplatform_getline(data_file,file_data_line);
226 if (file_data_line.substr(0,1) == ">")
228 else sequence_raw += file_data_line;
231 // Lastly, if subselection of the sequence was specified we keep cut out
232 // and only keep that part
233 // end_index = 0 means no end was specified, so cut to the end
235 end_index = sequence_raw.size();
237 // sequence filtering for upcasing agctn and convert non AGCTN to N
238 if (end_index-start_index <= 0) {
239 std::string msg("The selected sequence appears to be empty");
240 throw sequence_empty_error(msg);
242 set_filtered_sequence(sequence_raw, start_index, end_index-start_index);
244 std::string errormsg("There were no fasta sequences");
245 throw sequence_empty_file_error(errormsg);
249 void Sequence::set_filtered_sequence(const std::string &old_seq,
250 std::string::size_type start,
251 std::string::size_type count)
253 char conversionTable[257];
256 count = old_seq.size() - start;
257 boost::shared_ptr<seq_string> new_seq(new seq_string);
258 new_seq->reserve(count);
260 // Make a conversion table
262 // everything we don't specify below will become 'N'
263 for(int table_i=0; table_i < 256; table_i++)
265 conversionTable[table_i] = 'N';
267 // add end of string character for printing out table for testing purposes
268 conversionTable[256] = '\0';
270 // we want these to map to themselves - ie not to change
271 conversionTable[(int)'A'] = 'A';
272 conversionTable[(int)'T'] = 'T';
273 conversionTable[(int)'G'] = 'G';
274 conversionTable[(int)'C'] = 'C';
276 conversionTable[(int)'a'] = 'A';
277 conversionTable[(int)'t'] = 'T';
278 conversionTable[(int)'g'] = 'G';
279 conversionTable[(int)'c'] = 'C';
281 // finally, the actual conversion loop
282 for(std::string::size_type seq_index = 0; seq_index < count; seq_index++)
284 new_seq->append(1, conversionTable[ (int)old_seq[seq_index+start]]);
290 Sequence::load_annot(fs::path file_path, int start_index, int end_index)
292 fs::fstream data_stream(file_path, std::ios::in);
295 throw mussa_load_error("Sequence File: " + file_path.string() + " not found");
298 // so i should probably be passing the parse function some iterators
299 // but the annotations files are (currently) small, so i think i can
300 // get away with loading the whole file into memory
303 while(data_stream.good()) {
309 parse_annot(data, start_index, end_index);
312 /* If this works, yikes, this is some brain hurting code.
314 * what's going on is that when pb_annot is instantiated it stores references
315 * to begin, end, name, type, declared in the parse function, then
316 * when operator() is called it grabs values from those references
317 * and uses that to instantiate an annot object and append that to our
320 * This weirdness is because the spirit library requires that actions
321 * conform to a specific prototype operator()(IteratorT, IteratorT)
322 * which doesn't provide any useful opportunity for me to actually
323 * grab the results of our parsing.
325 * so I instantiate this structure in order to have a place to grab
329 struct push_back_annot {
330 std::list<annot>& annot_list;
336 push_back_annot(std::list<annot>& annot_list_,
341 : annot_list(annot_list_),
349 void operator()(std::string::const_iterator,
350 std::string::const_iterator) const
352 //std::cout << "adding annot: " << begin << "|" << end << "|" << name << "|" << type << std::endl;
353 annot_list.push_back(annot(begin, end, name, type));
357 struct push_back_seq {
358 std::list<Sequence>& seq_list;
362 push_back_seq(std::list<Sequence>& seq_list_,
365 : seq_list(seq_list_),
371 void operator()(std::string::const_iterator,
372 std::string::const_iterator) const
374 // filter out newlines from our sequence
376 for(std::string::const_iterator seq_i = seq.begin();
380 if (*seq_i != '\015' && *seq_i != '\012') new_seq += *seq_i;
382 //std::cout << "adding seq: " << name << " " << new_seq << std::endl;
385 s.set_fasta_header(name);
386 seq_list.push_back(s);
391 Sequence::parse_annot(std::string data, int start_index, int end_index)
398 std::list<Sequence> query_seqs;
400 bool status = spirit::parse(data.begin(), data.end(),
407 )[spirit::assign_a(species)] >>
411 ( // ignore html tags
412 *(spirit::space_p) >>
414 +(~spirit::ch_p('>')) >>
419 ( // parse an absolute location name
420 (spirit::uint_p[spirit::assign_a(start)] >>
422 spirit::uint_p[spirit::assign_a(end)] >>
427 )[spirit::assign_a(name)] >>
434 )[spirit::assign_a(type)]
436 // to understand how this group gets set
437 // read the comment above struct push_back_annot
438 )[push_back_annot(annots, start, end, type, name)]
440 ((spirit::ch_p('>')|spirit::str_p(">")) >>
441 (*(spirit::print_p))[spirit::assign_a(name)] >>
443 (+(spirit::chset<>(nucleic_iupac_alphabet.c_str())))[spirit::assign_a(seq)]
444 )[push_back_seq(query_seqs, name, seq)]
450 spirit::space_p*/).full;
452 // go seearch for query sequences
453 find_sequences(query_seqs.begin(), query_seqs.end());
457 void Sequence::add_annotation(const annot& a)
462 const std::list<annot>& Sequence::annotations() const
468 Sequence::subseq(int start, int count) const
470 // there might be an off by one error with start+count > size()
471 if ( count == npos || start+count > size()) {
472 count = size()-start;
474 Sequence new_seq(seq->substr(start, count));
475 new_seq.set_fasta_header(get_fasta_header());
476 new_seq.set_species(get_species());
478 new_seq.motif_list = motif_list;
479 // attempt to copy & reannotate position based annotations
480 int end = start+count;
482 for(std::list<annot>::const_iterator annot_i = annots.begin();
483 annot_i != annots.end();
486 int annot_begin= annot_i->begin;
487 int annot_end = annot_i->end;
489 if (annot_begin < end) {
490 if (annot_begin >= start) {
491 annot_begin -= start;
496 if (annot_end < end) {
502 annot new_annot(annot_begin, annot_end, annot_i->type, annot_i->name);
503 new_seq.annots.push_back(new_annot);
511 Sequence::rev_comp() const
513 std::string rev_comp;
514 char conversionTable[257];
515 int seq_i, table_i, len;
518 rev_comp.reserve(len);
519 // make a conversion table
520 // init all parts of conversion table to '~' character
521 // '~' I doubt will ever appear in a sequence file (jeez, I hope)
522 // and may the fleas of 1000 camels infest the genitals of any biologist (and
523 // seven generations of their progeny) who decides to make it mean
524 // something special!!!
525 // PS - double the curse for any smartass non-biologist who tries it as well
526 for(table_i=0; table_i < 256; table_i++)
528 conversionTable[table_i] = '~';
530 // add end of string character for printing out table for testing purposes
531 conversionTable[256] = '\0';
533 // add in the characters for the bases we want to convert
534 conversionTable[(int)'A'] = 'T';
535 conversionTable[(int)'T'] = 'A';
536 conversionTable[(int)'G'] = 'C';
537 conversionTable[(int)'C'] = 'G';
538 conversionTable[(int)'N'] = 'N';
540 // finally, the actual conversion loop
541 for(seq_i = len - 1; seq_i >= 0; seq_i--)
543 table_i = (int) seq->at(seq_i);
544 rev_comp += conversionTable[table_i];
550 void Sequence::set_fasta_header(std::string header_)
555 void Sequence::set_species(const std::string& name)
560 std::string Sequence::get_species() const
567 Sequence::get_fasta_header() const
573 Sequence::get_name() const
575 if (header.size() > 0)
577 else if (species.size() > 0)
583 void Sequence::set_sequence(const std::string& s)
585 set_filtered_sequence(s);
588 std::string Sequence::get_sequence() const
593 Sequence::const_reference Sequence::operator[](Sequence::size_type i) const
598 Sequence::const_reference Sequence::at(Sequence::size_type i) const
613 const char *Sequence::c_str() const
618 Sequence::const_iterator Sequence::begin() const
623 return Sequence::const_iterator(0);
626 Sequence::const_iterator Sequence::end() const
631 return Sequence::const_iterator(0);
634 bool Sequence::empty() const
642 Sequence::size_type Sequence::size() const
650 Sequence::size_type Sequence::length() const
655 Sequence::save(fs::fstream &save_file)
656 //std::string save_file_path)
659 std::list<annot>::iterator annots_i;
661 // not sure why, or if i'm doing something wrong, but can't seem to pass
662 // file pointers down to this method from the mussa control class
663 // so each call to save a sequence appends to the file started by mussa_class
664 //save_file.open(save_file_path.c_str(), std::ios::app);
666 save_file << "<Sequence>" << std::endl;
667 save_file << *this << std::endl;
668 save_file << "</Sequence>" << std::endl;
670 save_file << "<Annotations>" << std::endl;
671 save_file << species << std::endl;
672 for (annots_i = annots.begin(); annots_i != annots.end(); ++annots_i)
674 save_file << annots_i->begin << " " << annots_i->end << " " ;
675 save_file << annots_i->name << " " << annots_i->type << std::endl;
677 save_file << "</Annotations>" << std::endl;
682 Sequence::load_museq(fs::path load_file_path, int seq_num)
684 fs::fstream load_file;
685 std::string file_data_line;
688 std::string::size_type space_split_i;
689 std::string annot_value;
692 load_file.open(load_file_path, std::ios::in);
695 // search for the seq_num-th sequence
696 while ( (!load_file.eof()) && (seq_counter < seq_num) )
698 getline(load_file,file_data_line);
699 if (file_data_line == "<Sequence>")
702 getline(load_file, file_data_line);
703 // looks like the sequence is written as a single line
704 set_filtered_sequence(file_data_line);
705 getline(load_file, file_data_line);
706 getline(load_file, file_data_line);
707 if (file_data_line == "<Annotations>")
709 getline(load_file, file_data_line);
710 species = file_data_line;
711 while ( (!load_file.eof()) && (file_data_line != "</Annotations>") )
713 getline(load_file,file_data_line);
714 if ((file_data_line != "") && (file_data_line != "</Annotations>"))
716 // need to get 4 values...almost same code 4 times...
717 // get annot start index
718 space_split_i = file_data_line.find(" ");
719 annot_value = file_data_line.substr(0,space_split_i);
720 an_annot.begin = atoi (annot_value.c_str());
721 file_data_line = file_data_line.substr(space_split_i+1);
722 // get annot end index
723 space_split_i = file_data_line.find(" ");
724 annot_value = file_data_line.substr(0,space_split_i);
725 an_annot.end = atoi (annot_value.c_str());
727 if (space_split_i == std::string::npos) // no entry for type or name
729 std::cout << "seq, annots - no type or name\n";
733 else // else get annot type
735 file_data_line = file_data_line.substr(space_split_i+1);
736 space_split_i = file_data_line.find(" ");
737 annot_value = file_data_line.substr(0,space_split_i);
738 an_annot.type = annot_value;
739 if (space_split_i == std::string::npos) // no entry for name
741 std::cout << "seq, annots - no name\n";
744 else // get annot name
746 file_data_line = file_data_line.substr(space_split_i+1);
747 space_split_i = file_data_line.find(" ");
748 annot_value = file_data_line.substr(0,space_split_i);
749 an_annot.type = annot_value;
752 annots.push_back(an_annot); // don't forget to actually add the annot
754 //std::cout << "seq, annots: " << an_annot.start << ", " << an_annot.end
755 // << "-->" << an_annot.type << "::" << an_annot.name << std::endl;
762 Sequence::rc_motif(std::string a_motif) const
764 std::string rev_comp;
765 char conversionTable[257];
766 int seq_i, table_i, len;
768 len = a_motif.length();
769 rev_comp.reserve(len);
771 for(table_i=0; table_i < 256; table_i++)
773 conversionTable[table_i] = '~';
775 // add end of std::string character for printing out table for testing purposes
776 conversionTable[256] = '\0';
778 // add in the characters for the bases we want to convert (IUPAC)
779 conversionTable[(int)'A'] = 'T';
780 conversionTable[(int)'T'] = 'A';
781 conversionTable[(int)'G'] = 'C';
782 conversionTable[(int)'C'] = 'G';
783 conversionTable[(int)'N'] = 'N';
784 conversionTable[(int)'M'] = 'K';
785 conversionTable[(int)'R'] = 'Y';
786 conversionTable[(int)'W'] = 'W';
787 conversionTable[(int)'S'] = 'S';
788 conversionTable[(int)'Y'] = 'R';
789 conversionTable[(int)'K'] = 'M';
790 conversionTable[(int)'V'] = 'B';
791 conversionTable[(int)'H'] = 'D';
792 conversionTable[(int)'D'] = 'H';
793 conversionTable[(int)'B'] = 'V';
795 // finally, the actual conversion loop
796 for(seq_i = len - 1; seq_i >= 0; seq_i--)
798 //std::cout << "** i = " << seq_i << " bp = " <<
799 table_i = (int) a_motif[seq_i];
800 rev_comp += conversionTable[table_i];
803 //std::cout << "seq: " << a_motif << std::endl;
804 //std::cout << "rc: " << rev_comp << std::endl;
810 Sequence::motif_normalize(const std::string& a_motif)
812 std::string valid_motif;
815 len = a_motif.length();
816 valid_motif.reserve(len);
818 // this just upcases IUPAC symbols. Eventually should return an error if non IUPAC is present.
819 // current nonIUPAC symbols are omitted, which is not reported atm
820 for(seq_i = 0; seq_i < len; seq_i++)
822 if ((a_motif[seq_i] == 'a') || (a_motif[seq_i] == 'A'))
824 else if ((a_motif[seq_i] == 't') || (a_motif[seq_i] == 'T'))
826 else if ((a_motif[seq_i] == 'g') || (a_motif[seq_i] == 'G'))
828 else if ((a_motif[seq_i] == 'c') || (a_motif[seq_i] == 'C'))
830 else if ((a_motif[seq_i] == 'n') || (a_motif[seq_i] == 'N'))
832 else if ((a_motif[seq_i] == 'm') || (a_motif[seq_i] == 'M'))
834 else if ((a_motif[seq_i] == 'r') || (a_motif[seq_i] == 'R'))
836 else if ((a_motif[seq_i] == 'w') || (a_motif[seq_i] == 'W'))
838 else if ((a_motif[seq_i] == 's') || (a_motif[seq_i] == 'S'))
840 else if ((a_motif[seq_i] == 'y') || (a_motif[seq_i] == 'Y'))
842 else if ((a_motif[seq_i] == 'k') || (a_motif[seq_i] == 'K'))
844 else if ((a_motif[seq_i] == 'v') || (a_motif[seq_i] == 'V'))
846 else if ((a_motif[seq_i] == 'h') || (a_motif[seq_i] == 'H'))
848 else if ((a_motif[seq_i] == 'd') || (a_motif[seq_i] == 'D'))
850 else if ((a_motif[seq_i] == 'b') || (a_motif[seq_i] == 'B'))
853 std::string msg = "Letter ";
854 msg += a_motif[seq_i];
855 msg += " is not a valid IUPAC symbol";
856 throw motif_normalize_error(msg);
859 //std::cout << "valid_motif is: " << valid_motif << std::endl;
863 void Sequence::add_motif(const Sequence& a_motif)
865 std::vector<int> motif_starts = find_motif(a_motif);
867 for(std::vector<int>::iterator motif_start_i = motif_starts.begin();
868 motif_start_i != motif_starts.end();
871 motif_list.push_back(motif(*motif_start_i, a_motif.get_sequence()));
875 void Sequence::clear_motifs()
880 const std::list<motif>& Sequence::motifs() const
886 Sequence::find_motif(const std::string& a_motif) const
888 std::vector<int> motif_match_starts;
889 std::string norm_motif_rc;
891 motif_match_starts.clear();
893 //std::cout << "motif is: " << a_motif << std::endl;
894 std::string norm_motif = motif_normalize(a_motif);
895 //std::cout << "motif is: " << a_motif << std::endl;
897 if (norm_motif.size() > 0)
899 //std::cout << "Sequence: none blank motif\n";
900 motif_scan(norm_motif, &motif_match_starts);
902 norm_motif_rc = rc_motif(a_motif);
903 // make sure not to do search again if it is a palindrome
904 if (norm_motif_rc != norm_motif) {
905 motif_scan(norm_motif_rc, &motif_match_starts);
908 return motif_match_starts;
912 Sequence::find_motif(const Sequence& a_motif) const
914 return find_motif(a_motif.get_sequence());
918 Sequence::motif_scan(std::string a_motif, std::vector<int> * motif_match_starts) const
920 std::string::const_iterator seq_c = seq->begin();
921 std::string::size_type seq_i;
922 int motif_i, motif_len;
924 //std::cout << "Sequence: motif, seq len = " << sequence.length() << std::endl;
925 motif_len = a_motif.length();
927 //std::cout << "motif_length: " << motif_len << std::endl;
928 //std::cout << "RAAARRRRR\n";
932 //std::cout << "motif: " << a_motif << std::endl;
934 //std::cout << "Sequence: motif, length= " << length << std::endl;
936 while (seq_i < length())
938 //std::cout << seq_c[seq_i];
939 //std::cout << seq_c[seq_i] << "?" << a_motif[motif_i] << ":" << motif_i << " ";
940 // this is pretty much a straight translation of Nora's python code
941 // to match iupac letter codes
942 if (a_motif[motif_i] =='N')
944 else if (a_motif[motif_i] == seq_c[seq_i])
946 else if ((a_motif[motif_i] =='M') &&
947 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C')))
949 else if ((a_motif[motif_i] =='R') &&
950 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G')))
952 else if ((a_motif[motif_i] =='W') &&
953 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='T')))
955 else if ((a_motif[motif_i] =='S') &&
956 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G')))
958 else if ((a_motif[motif_i] =='Y') &&
959 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='T')))
961 else if ((a_motif[motif_i] =='K') &&
962 ((seq_c[seq_i]=='G') || (seq_c[seq_i]=='T')))
964 else if ((a_motif[motif_i] =='V') &&
965 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
966 (seq_c[seq_i]=='G')))
968 else if ((a_motif[seq_i] =='H') &&
969 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='C') ||
970 (seq_c[seq_i]=='T')))
972 else if ((a_motif[motif_i] =='D') &&
973 ((seq_c[seq_i]=='A') || (seq_c[seq_i]=='G') ||
974 (seq_c[seq_i]=='T')))
976 else if ((a_motif[motif_i] =='B') &&
977 ((seq_c[seq_i]=='C') || (seq_c[seq_i]=='G') ||
978 (seq_c[seq_i]=='T')))
987 // end Nora stuff, now we see if a match is found this pass
988 if (motif_i == motif_len)
992 motif_match_starts->push_back(seq_i - motif_len + 1);
998 //std::cout << std::endl;
1001 void Sequence::add_string_annotation(std::string a_seq,
1004 std::vector<int> seq_starts = find_motif(a_seq);
1006 //std::cout << "searching for " << a_seq << " found " << seq_starts.size() << std::endl;
1008 for(std::vector<int>::iterator seq_start_i = seq_starts.begin();
1009 seq_start_i != seq_starts.end();
1012 annots.push_back(annot(*seq_start_i,
1013 *seq_start_i+a_seq.size(),
1019 void Sequence::find_sequences(std::list<Sequence>::iterator start,
1020 std::list<Sequence>::iterator end)
1022 while (start != end) {
1023 add_string_annotation(start->get_sequence(), start->get_fasta_header());
1029 std::ostream& operator<<(std::ostream& out, const Sequence& s)
1031 for(Sequence::const_iterator s_i = s.begin(); s_i != s.end(); ++s_i) {
1037 bool operator<(const Sequence& x, const Sequence& y)
1039 Sequence::const_iterator x_i = x.begin();
1040 Sequence::const_iterator y_i = y.begin();
1043 if( x_i == x.end() and y_i == y.end() ) {
1045 } else if ( x_i == x.end() ) {
1047 } else if ( y_i == y.end() ) {
1049 } else if ( (*x_i) < (*y_i)) {
1051 } else if ( (*x_i) > (*y_i) ) {
1060 bool operator==(const Sequence& x, const Sequence& y)
1062 if (x.empty() and y.empty()) {
1063 // if there's no sequence in either sequence structure, they're equal
1065 } else if (x.empty() or y.empty()) {
1066 // if we fail the first test, and we discover one is empty,
1067 // we know they can't be equal. (and we need to do this
1068 // to prevent dereferencing an empty pointer)
1070 } else if ( *(x.seq) == *(y.seq)) {
1071 // and x.annots == y.annots and x.motif_list == y.motif_list) {