Consensus.hpp

/*
    Copyright 2008-2009 Stéphane De Mita, Mathieu Siol

    This file is part of the EggLib library.

    EggLib is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation, either version 3 of the License, or
    (at your option) any later version.

    EggLib is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with EggLib.  If not, see <http://www.gnu.org/licenses/>.
*/

#ifndef EGGLIB_CONSENSUS_HPP
#define EGGLIB_CONSENSUS_HPP

#include "Align.hpp"
#include <sstream>
#include <string>
#include <vector>

namespace egglib {

    class Consensus {
       public:
          Consensus();
          
          virtual ~Consensus() {}
          
          void setMissing(char);
          
          void setDisagreement(char);
          
          bool check_sequences(Align& align);
          
          Align consensus(Align& align, char separator='_', bool rigorous=true);

          const std::vector<std::string>& firstSequenceNames();

          const std::vector<std::string>& secondSequenceNames();
          
          const std::vector<std::string>& roots();
          
          const std::vector<int>& consistentPositions();

          const std::vector<int>& complementaryPositions();

          const std::vector<int>& uninformativePositions();

          const std::vector<int>& ambiguousPositions();

          const std::vector<int>& atLeastPartiallyResolvedAmbiguities();

          const std::vector<std::vector<int> >& inconsistentPositions();


       private:

            Consensus(const Consensus& source) { }
            
            Consensus& operator=(const Consensus& source) { return *this; }
            
          // A private helper
          class PairwiseConsensus;
         
          // report data
          std::vector<std::string> t_firstSequenceNames;
          std::vector<std::string> t_secondSequenceNames;
          std::vector<std::string> t_roots;
          std::vector<int> t_consistentPositions;
          std::vector<int> t_complementaryPositions;
          std::vector<int> t_uninformativePositions;
          std::vector<int> t_ambiguousPositions;
          std::vector<int> t_atLeastPartiallyResolvedAmbiguities;
          std::vector<std::vector<int> > t_inconsistentPositions;

          // Code for missing data (usually ?)
          char MISSING;

          // Code for disgrement
          char DISAGREEMENT;

          // Helper class managing a single pair
          class PairwiseConsensus {
              public:
                 // Default object creation
                 PairwiseConsensus();
                 
                 // Object destruction
                 virtual ~PairwiseConsensus() {}
                 
                 // Usual object creation
                 PairwiseConsensus(std::string, std::string);
                 
                 // Fills an object created with the default constructor
                 void load(std::string,std::string);
                 
                 // Changes the MISSING character
                 void setUndeterminedCharacter(char);
                 
                 // Changes the DISAGREEMENT character
                 void setDisagreementCharacter(char);
                 
                 /* Uses the conservative mode of consensus
                  * 
                  * Tries to avoid to make decisions, and adds the
                  * character set by DISAGREEMENT upon inconsistencies
                  * 
                  * return The number of inconsistencies.
                  */
                 int generateSoftConsensus();
                 
                 /* Strict mode of consensus
                  * 
                  * The number of inconsistencies.
                  */
                 int generateHardConsensus();
                 
                 // Two fully resolved (including gap) and identical characters
                 int getConsistentPositions();

                 // One informative (including gap) and one missing
                 int getComplementaryPositions();

                 // None missing, but different and incompatible
                 int getInconsistentPositions();
                 
                 // Both are missing
                 int getUninformativePositions();
                 
                 // Both identical or one missing, but not fully resolved
                 int getAmbiguousPositions();
                 
                 // Different, not missing, complementary.
                 int getAtLeastPartiallyResolvedAmbiguities();
                 
                 // Accessor
                 int getThisInconsistentPosition(unsigned int);
                 
                 // Generates the consensus sequence
                 std::string getConsensus();

              private:
              
                inline bool isValid(char c) {
                    switch (c) {
                        case 'A': case 'a':
                        case 'C': case 'c':
                        case 'G': case 'g':
                        case 'T': case 't': 
                            return true;
                        default:
                            return false;
                    }
                }
              
                 // This initiates a series of embedded objects
                 void setCharacterContainers();
                 
                 // The first sequence
                 std::string seqA;
                 
                 // The second sequence
                 std::string seqB;
                 
                 // The resulting consensus
                 std::string cons;
                 
                 // The vecotr storing the inconsistent positions
                 std::vector<int> posIncons;
                 
                 // The length of the sequences
                 unsigned int ls;
                 
                 // Counter
                 int cntConsistentPositions;
                 
                 // Counter
                 int cntComplementaryPositions;
                 
                 // Counter
                 int cntAmbiguousPositions;
                 
                 // Counter
                 int cntInconsistentPositions;
                 
                 // Counter
                 int cntUninformativePositions;
                 
                 // Counter
                 int cntAtLeastPartiallyResolvedAmbiguities;

                 // Code for missing data (usually ?)
                 char MISSING;

                 // Code for disgrement
                 char DISAGREEMENT;
     
            public:
                // This class manages relationships different symbols
                class CharacterContainer {
                      public:
                         // Default value: @
                         CharacterContainer();
                         
                         // Initiates to a given symbol
                         CharacterContainer(const char&);
                         
                         // Assignment operator
                         CharacterContainer& operator=(const char&);
                         
                         // Sets the symbol
                         void setValue(char);
                         
                         // Set the descendants
                         void setSons(std::vector<CharacterContainer>);
                         
                         // Tests whether the symbol is the same
                         bool is(CharacterContainer);
                         
                         // Tests if the query is contained amongst the sons
                         bool has(CharacterContainer);
                         
                         // Tests if the query is contained amongst the sons
                         bool has(char);
                         
                         /* Tests whether the left character has the left one
                          * Should be called on the N object only.
                          */
                         char lhas(CharacterContainer,CharacterContainer);
                         
                         /* Creates the object with the proper sons
                          * Should be called on the N object only.
                          */
                         CharacterContainer init(char);
                         
                         // The symbol
                         char value;
                         
                         // The descendants
                         std::vector<CharacterContainer> sons;
                  };

             private:
                  // Symbol ?
                  CharacterContainer ccQ;
                  
                  // Symbol A
                  CharacterContainer ccA;
                  
                  // Symbol C
                  CharacterContainer ccC;
                  
                  // Symbol G
                  CharacterContainer ccG;
                  
                  // Symbol T
                  CharacterContainer ccT;
                  
                  // Symbol U
                  CharacterContainer ccU;
                  
                  // Symbol M
                  CharacterContainer ccM;
                  
                  // Symbol R
                  CharacterContainer ccR;
                  
                  // Symbol W
                  CharacterContainer ccW;
                  
                  // Symbol S
                  CharacterContainer ccS;
                  
                  // Symbol Y
                  CharacterContainer ccY;
                  
                  // Symbol K
                  CharacterContainer ccK;
                  
                  // Symbol B
                  CharacterContainer ccB;
                  
                  // Symbol D
                  CharacterContainer ccD;
                  
                  // Symbol H
                  CharacterContainer ccH;
                  
                  // Symbol V
                  CharacterContainer ccV;
                  
                  // Symbol N
                  CharacterContainer ccN;
                  
                  // Symbol -
                  CharacterContainer ccGAP;
          };
    };
}

#endif