Bpp/Seq/SequenceTools.cpp

Go to the documentation of this file.

//
// File: SequenceTools.cpp
// Authors: Guillaume Deuchst
//          Julien Dutheil
//          Sylvain Gaillard
// Created on: Tue Aug 21 2003
//

/*
   Copyright or © or Copr. Bio++ Development Team, (November 17, 2004)

   This software is a computer program whose purpose is to provide classes
   for sequences analysis.

   This software is governed by the CeCILL  license under French law and
   abiding by the rules of distribution of free software.  You can  use,
   modify and/ or redistribute the software under the terms of the CeCILL
   license as circulated by CEA, CNRS and INRIA at the following URL
   "http://www.cecill.info".

   As a counterpart to the access to the source code and  rights to copy,
   modify and redistribute granted by the license, users are provided only
   with a limited warranty  and the software's author,  the holder of the
   economic rights,  and the successive licensors  have only  limited
   liability.

   In this respect, the user's attention is drawn to the risks associated
   with loading,  using,  modifying and/or developing or reproducing the
   software by the user in light of its specific status of free software,
   that may mean  that it is complicated to manipulate,  and  that  also
   therefore means  that it is reserved for developers  and  experienced
   professionals having in-depth computer knowledge. Users are therefore
   encouraged to load and test the software's suitability as regards their
   requirements in conditions enabling the security of their systems and/or
   data to be ensured and,  more generally, to use and operate it in the
   same conditions as regards security.

   The fact that you are presently reading this means that you have had
   knowledge of the CeCILL license and that you accept its terms.
 */

#include "SequenceTools.h"

#include "Alphabet/AlphabetTools.h"
#include "StringSequenceTools.h"
#include <Bpp/Numeric/Matrix/Matrix.h>
#include <Bpp/Numeric/VectorTools.h>

using namespace bpp;

// From the STL:
#include <ctype.h>
#include <cmath>
#include <list>
#include <iostream>

using namespace std;

DNA SequenceTools::_DNA;
RNA SequenceTools::_RNA;
RNY SequenceTools::_RNY(_DNA);
NucleicAcidsReplication SequenceTools::_DNARep(&_DNA, &_DNA);
NucleicAcidsReplication SequenceTools::_RNARep(&_RNA, &_RNA);
NucleicAcidsReplication SequenceTools::_transc(&_DNA, &_RNA);

/******************************************************************************/

Sequence* SequenceTools::subseq(const Sequence& sequence, size_t begin, size_t end) throw (IndexOutOfBoundsException, Exception)
{
  // Checking interval
  if (end >= sequence.size())
    throw IndexOutOfBoundsException ("SequenceTools::subseq : Invalid upper bound", end, 0, sequence.size());
  if (end < begin)
    throw Exception ("SequenceTools::subseq : Invalid interval");

  // Copy sequence
  vector<int> temp(sequence.getContent());

  // Truncate sequence
  temp.erase(temp.begin() + end + 1, temp.end());
  temp.erase(temp.begin(), temp.begin() + begin);

  // New sequence creation
  return new BasicSequence(sequence.getName(), temp, sequence.getComments(), sequence.getAlphabet());
}

/******************************************************************************/

Sequence* SequenceTools::concatenate(const Sequence& seq1, const Sequence& seq2) throw (AlphabetMismatchException, Exception)
{
  // Sequence's alphabets matching verification
  if ((seq1.getAlphabet()->getAlphabetType()) != (seq2.getAlphabet()->getAlphabetType()))
    throw AlphabetMismatchException("SequenceTools::concatenate : Sequence's alphabets don't match ", seq1.getAlphabet(), seq2.getAlphabet());

  // Sequence's names matching verification
  if (seq1.getName() != seq2.getName())
    throw Exception ("SequenceTools::concatenate : Sequence's names don't match");

  // Concatenate sequences and send result
  vector<int> sequence  = seq1.getContent();
  vector<int> sequence2 = seq2.getContent();
  sequence.insert(sequence.end(), sequence2.begin(), sequence2.end());
  return new BasicSequence(seq1.getName(), sequence, seq1.getComments(), seq1.getAlphabet());
}

/******************************************************************************/

Sequence& SequenceTools::complement(Sequence& seq) throw (AlphabetException)
{
  // Alphabet type checking
  NucleicAcidsReplication* NAR;
  if (seq.getAlphabet()->getAlphabetType() == "DNA alphabet")
  {
    NAR = &_DNARep;
  }
  else if (seq.getAlphabet()->getAlphabetType() == "RNA alphabet")
  {
    NAR = &_RNARep;
  }
  else
  {
    throw AlphabetException("SequenceTools::complement: Sequence must be nucleic.", seq.getAlphabet());
  }
  for (size_t i = 0; i < seq.size(); i++)
  {
    seq.setElement(i, NAR->translate(seq.getValue(i)));
  }
  return seq;
}

/******************************************************************************/

Sequence* SequenceTools::getComplement(const Sequence& sequence) throw (AlphabetException)
{
  // Alphabet type checking
  NucleicAcidsReplication* NAR;
  if (sequence.getAlphabet()->getAlphabetType() == "DNA alphabet")
  {
    NAR = &_DNARep;
  }
  else if (sequence.getAlphabet()->getAlphabetType() == "RNA alphabet")
  {
    NAR = &_RNARep;
  }
  else
  {
    throw AlphabetException ("SequenceTools::getComplement: Sequence must be nucleic.", sequence.getAlphabet());
  }

  return NAR->translate(sequence);
}

/******************************************************************************/

Sequence* SequenceTools::transcript(const Sequence& sequence) throw (AlphabetException)
{
  // Alphabet type checking
  if (sequence.getAlphabet()->getAlphabetType() != "DNA alphabet")
  {
    throw AlphabetException ("SequenceTools::transcript : Sequence must be DNA", sequence.getAlphabet());
  }

  return _transc.translate(sequence);
}

/******************************************************************************/

Sequence* SequenceTools::reverseTranscript(const Sequence& sequence) throw (AlphabetException)
{
  // Alphabet type checking
  if (sequence.getAlphabet()->getAlphabetType() != "RNA alphabet")
  {
    throw AlphabetException ("SequenceTools::reverseTranscript : Sequence must be RNA", sequence.getAlphabet());
  }

  return _transc.reverse(sequence);
}

/******************************************************************************/

Sequence& SequenceTools::invert(Sequence& seq)
{
  size_t seq_size = seq.size(); // store seq size for efficiency
  unsigned int tmp_state = 0; // to store one state when swapping positions
  size_t j = seq_size; // symetric position iterator from sequence end
  for (size_t i = 0; i < seq_size / 2; i++)
  {
    j = seq_size - 1 - i;
    tmp_state = seq.getValue(i);
    seq.setElement(i, seq.getValue(j));
    seq.setElement(j, tmp_state);
  }
  return seq;
}

/******************************************************************************/

Sequence* SequenceTools::getInvert(const Sequence& sequence)
{
  Sequence* iSeq = sequence.clone();
  invert(*iSeq);
  return iSeq;
}

/******************************************************************************/

Sequence& SequenceTools::invertComplement(Sequence& seq)
{
  // Alphabet type checking
  NucleicAcidsReplication* NAR;
  if (seq.getAlphabet()->getAlphabetType() == "DNA alphabet")
  {
    NAR = &_DNARep;
  }
  else if (seq.getAlphabet()->getAlphabetType() == "RNA alphabet")
  {
    NAR = &_RNARep;
  }
  else
  {
    throw AlphabetException("SequenceTools::complement: Sequence must be nucleic.", seq.getAlphabet());
  }
  // for (size_t i = 0 ; i < seq.size() ; i++) {
  //  seq.setElement(i, NAR->translate(seq.getValue(i)));
  // }
  size_t seq_size = seq.size(); // store seq size for efficiency
  int tmp_state = 0; // to store one state when swapping positions
  size_t j = seq_size; // symetric position iterator from sequence end
  for (size_t i = 0; i < seq_size / 2; i++)
  {
    j = seq_size - 1 - i;
    tmp_state = seq.getValue(i);
    seq.setElement(i, NAR->translate(seq.getValue(j)));
    seq.setElement(j, NAR->translate(tmp_state));
  }
  if (seq_size % 2)   // treate the state in the middle of odd sequences
  {
    seq.setElement(seq_size / 2, NAR->translate(seq.getValue(seq_size / 2)));
  }
  return seq;
}

/******************************************************************************/

double SequenceTools::getPercentIdentity(const Sequence& seq1, const Sequence& seq2, bool ignoreGaps) throw (AlphabetMismatchException, SequenceNotAlignedException)
{
  if (seq1.getAlphabet()->getAlphabetType() != seq2.getAlphabet()->getAlphabetType())
    throw AlphabetMismatchException("SequenceTools::getPercentIdentity", seq1.getAlphabet(), seq2.getAlphabet());
  if (seq1.size() != seq2.size())
    throw SequenceNotAlignedException("SequenceTools::getPercentIdentity", &seq2);
  int gap = seq1.getAlphabet()->getGapCharacterCode();
  size_t id = 0;
  size_t tot = 0;
  for (size_t i = 0; i < seq1.size(); i++)
  {
    int x = seq1.getValue(i);
    int y = seq2.getValue(i);
    if (ignoreGaps)
    {
      if (x != gap && y != gap)
      {
        tot++;
        if (x == y)
          id++;
      }
    }
    else
    {
      tot++;
      if (x == y)
        id++;
    }
  }
  return static_cast<double>(id) / static_cast<double>(tot) * 100.;
}

/******************************************************************************/

size_t SequenceTools::getNumberOfSites(const Sequence& seq)
{
  size_t count = 0;
  const Alphabet* alpha = seq.getAlphabet();
  for (size_t i = 0; i < seq.size(); i++)
  {
    if (!alpha->isGap(seq[i]))
      count++;
  }
  return count;
}

/******************************************************************************/

size_t SequenceTools::getNumberOfCompleteSites(const Sequence& seq)
{
  size_t count = 0;
  const Alphabet* alpha = seq.getAlphabet();
  for (size_t i = 0; i < seq.size(); i++)
  {
    if (!alpha->isGap(seq[i]) && !alpha->isUnresolved(seq[i]))
      count++;
  }
  return count;
}

/******************************************************************************/

size_t SequenceTools::getNumberOfUnresolvedSites(const Sequence& seq)
{
  size_t count = 0;
  const Alphabet* alpha = seq.getAlphabet();
  for (size_t i = 0; i < seq.size(); i++)
  {
    if (alpha->isUnresolved(seq[i]))
      count++;
  }
  return count;
}

/******************************************************************************/

Sequence* SequenceTools::getSequenceWithoutGaps(const Sequence& seq)
{
  const Alphabet* alpha = seq.getAlphabet();
  vector<int> content;
  for (size_t i = 0; i < seq.size(); i++)
  {
    if (!alpha->isGap(seq[i]))
      content.push_back(seq[i]);
  }
  Sequence* newSeq = dynamic_cast<Sequence*>(seq.clone());
  newSeq->setContent(content);
  return newSeq;
}

/******************************************************************************/

void SequenceTools::removeGaps(Sequence& seq)
{
  const Alphabet* alpha = seq.getAlphabet();
  for (size_t i = seq.size(); i > 0; --i)
  {
    if (alpha->isGap(seq[i - 1]))
      seq.deleteElement(i - 1);
  }
}

/******************************************************************************/

Sequence* SequenceTools::getSequenceWithoutStops(const Sequence& seq) throw (Exception)
{
  const CodonAlphabet* calpha = dynamic_cast<const CodonAlphabet*>(seq.getAlphabet());
  if (!calpha)
    throw Exception("SequenceTools::getSequenceWithoutStops. Input sequence should have a codon alphabet.");
  vector<int> content;
  for (size_t i = 0; i < seq.size(); i++)
  {
    if (!calpha->isStop(seq[i]))
      content.push_back(seq[i]);
  }
  Sequence* newSeq = dynamic_cast<Sequence*>(seq.clone());
  newSeq->setContent(content);
  return newSeq;
}

/******************************************************************************/

void SequenceTools::removeStops(Sequence& seq) throw (Exception)
{
  const CodonAlphabet* calpha = dynamic_cast<const CodonAlphabet*>(seq.getAlphabet());
  if (!calpha)
    throw Exception("SequenceTools::removeStops. Input sequence should have a codon alphabet.");
  for (size_t i = seq.size(); i > 0; --i)
  {
    if (calpha->isStop(seq[i - 1]))
      seq.deleteElement(i - 1);
  }
}

/******************************************************************************/

void SequenceTools::replaceStopsWithGaps(Sequence& seq) throw (Exception)
{
  const CodonAlphabet* calpha = dynamic_cast<const CodonAlphabet*>(seq.getAlphabet());
  if (!calpha)
    throw Exception("SequenceTools::replaceStopsWithGaps. Input sequence should have a codon alphabet.");
  int gap = calpha->getGapCharacterCode();
  for (size_t i = 0; i < seq.size(); ++i)
  {
    if (calpha->isStop(seq[i]))
      seq.setElement(i, gap);
  }
}

/******************************************************************************/

BowkerTest* SequenceTools::bowkerTest(const Sequence& seq1, const Sequence& seq2)
throw (SequenceNotAlignedException)
{
  if (seq1.size() != seq2.size())
    throw SequenceNotAlignedException("SequenceTools::bowkerTest.", &seq2);
  size_t n = seq1.size();
  const Alphabet* alpha = seq1.getAlphabet();
  unsigned int r = alpha->getSize();

  // Compute contingency table:
  RowMatrix<double> array(r, r);
  int x, y;
  for (size_t i = 0; i < n; i++)
  {
    x = seq1[i];
    y = seq2[i];
    if (!alpha->isGap(x) && !alpha->isUnresolved(x)
        && !alpha->isGap(y) && !alpha->isUnresolved(y))
    {
      array(static_cast<size_t>(x), static_cast<size_t>(y))++;
    }
  }

  // Compute Bowker's statistic:
  double sb2 = 0, nij, nji;
  for (unsigned int i = 1; i < r; ++i)
  {
    for (unsigned int j = 0; j < i; ++j)
    {
      nij = array(i, j);
      nji = array(j, i);
      if (nij != 0 || nji != 0)
      {
        sb2 += pow(nij - nji, 2) / (nij + nji);
      }
      // Else: we should display a warning there.
    }
  }

  // Compute p-value:
  double pvalue = 1. - RandomTools::pChisq(sb2, (r - 1) * r / 2);

  // Return results:
  BowkerTest* test = new BowkerTest();
  test->setStatistic(sb2);
  test->setPValue(pvalue);
  return test;
}

/******************************************************************************/

void SequenceTools::getPutativeHaplotypes(const Sequence& seq, std::vector<Sequence*>& hap, unsigned int level)
{
  vector< vector< int > > states(seq.size());
  list<Sequence*> t_hap;
  const Alphabet* alpha = seq.getAlphabet();
  unsigned int hap_count = 1;
  // Vector of available states at each position
  for (size_t i = 0; i < seq.size(); i++)
  {
    vector<int> st = alpha->getAlias(seq[i]);
    if (!st.size())
    {
      st.push_back(alpha->getGapCharacterCode());
    }
    if (st.size() <= level)
    {
      states[i] = st;
    }
    else
    {
      states[i] = vector<int>(1, seq[i]);
    }
  }
  // Combinatorial haplotypes building (the use of tree may be more accurate)
  t_hap.push_back(new BasicSequence(seq.getName() + "_hap" + TextTools::toString(hap_count++), "", alpha));
  for (size_t i = 0; i < states.size(); i++)
  {
    for (list<Sequence*>::iterator it = t_hap.begin(); it != t_hap.end(); it++)
    {
      for (unsigned int j = 0; j < states[i].size(); j++)
      {
        Sequence* tmp_seq = new BasicSequence(seq.getName() + "_hap", (**it).getContent(), alpha);
        if (j < states[i].size() - 1)
        {
          tmp_seq->setName(tmp_seq->getName() + TextTools::toString(hap_count++));
          tmp_seq->addElement(states[i][j]);
          t_hap.insert(it, tmp_seq);
        }
        else
        {
          (**it).addElement(states[i][j]);
        }
      }
    }
  }
  for (list<Sequence*>::reverse_iterator it = t_hap.rbegin(); it != t_hap.rend(); it++)
  {
    hap.push_back(*it);
  }
}

/******************************************************************************/

Sequence* SequenceTools::combineSequences(const Sequence& s1, const Sequence& s2) throw (AlphabetMismatchException)
{
  if (s1.getAlphabet()->getAlphabetType() != s2.getAlphabet()->getAlphabetType())
  {
    throw AlphabetMismatchException("SequenceTools::combineSequences(const Sequence& s1, const Sequence& s2): s1 and s2 don't have same Alphabet.", s1.getAlphabet(), s2.getAlphabet());
  }
  const Alphabet* alpha = s1.getAlphabet();
  vector<int> st;
  vector<int> seq;
  size_t length = NumTools::max(s1.size(), s2.size());
  for (size_t i = 0; i < length; i++)
  {
    if (i < s1.size())
      st.push_back(s1.getValue(i));
    if (i < s2.size())
      st.push_back(s2.getValue(i));
    seq.push_back(alpha->getGeneric(st));
    st.clear();
  }
  Sequence* s = new BasicSequence(s1.getName() + "+" + s2.getName(), seq, alpha);
  return s;
}

/******************************************************************************/

Sequence* SequenceTools::subtractHaplotype(const Sequence& s, const Sequence& h, string name, unsigned int level) throw (SequenceNotAlignedException)
{
  const Alphabet* alpha = s.getAlphabet();
  if (name.size() == 0)
    name = s.getName() + "_haplotype";
  string seq;
  if (s.size() != h.size())
    throw SequenceNotAlignedException("SequenceTools::subtractHaplotype: haplotype must be aligned with the sequence.", &h);
  for (unsigned int i = 0; i < s.size(); ++i)
  {
    string c;
    vector<int> nucs = alpha->getAlias(s.getValue(i));
    if (nucs.size() > 1)
    {
      remove(nucs.begin(), nucs.end(), h.getValue(i));
      nucs = vector<int>(nucs.begin(), nucs.end() - 1);
    }
    else
    {
      nucs = vector<int>(nucs.begin(), nucs.end());
    }
    c = alpha->intToChar(alpha->getGeneric(nucs));
    if (level <= nucs.size() && (alpha->isUnresolved(s.getValue(i)) || alpha->isUnresolved(h.getValue(i))))
    {
      c = alpha->intToChar(alpha->getUnknownCharacterCode());
    }
    seq += c;
  }
  Sequence* hap = new BasicSequence(name, seq, alpha);
  return hap;
}

/******************************************************************************/

Sequence* SequenceTools::RNYslice(const Sequence& seq, int ph) throw (AlphabetException)
{
  // Alphabet type checking
  if (seq.getAlphabet()->getAlphabetType() != "DNA alphabet")
  {
    throw AlphabetException ("SequenceTools::transcript : Sequence must be DNA", seq.getAlphabet());
  }

  if ((ph < 1) || (ph > 3))
    throw Exception("Bad phase for RNYSlice: " + TextTools::toString(ph) + ". Should be between 1 and 3.");

  size_t s = seq.size();
  size_t n = (s - ph + 3) / 3;

  vector<int> content(n);

  int tir = seq.getAlphabet()->getGapCharacterCode();
  size_t j;

  for (size_t i = 0; i < n; i++)
  {
    j = i * 3 + ph - 1;

    if (j == 0)
      content[i] = _RNY.getRNY(tir, seq[0], seq[1], *seq.getAlphabet());
    else
    {
      if (j == s - 1)
        content[i] = _RNY.getRNY(seq[j - 1], seq[j], tir, *seq.getAlphabet());
      else
        content[i] = _RNY.getRNY(seq[j - 1], seq[j], seq[j + 1], *seq.getAlphabet());
    }
  }

  // New sequence creating, and sense reversing
  Sequence* sq = new BasicSequence(seq.getName(), content,
                                   seq.getComments(), &_RNY);

  // Send result
  return sq;
}

Sequence* SequenceTools::RNYslice(const Sequence& seq) throw (AlphabetException)
{
  // Alphabet type checking
  if (seq.getAlphabet()->getAlphabetType() != "DNA alphabet")
  {
    throw AlphabetException ("SequenceTools::transcript : Sequence must be DNA", seq.getAlphabet());
  }

  size_t n = seq.size();

  vector<int> content(n);

  int tir = seq.getAlphabet()->getGapCharacterCode();

  if (seq.size() >= 2)
  {
    content[0] = _RNY.getRNY(tir, seq[0], seq[1], *seq.getAlphabet());

    for (unsigned int i = 1; i < n - 1; i++)
    {
      content[i] = _RNY.getRNY(seq[i - 1], seq[i], seq[i + 1],
                               *seq.getAlphabet());
    }

    content[n - 1] = _RNY.getRNY(seq[n - 2], seq[n - 1], tir, *seq.getAlphabet());
  }

  // New sequence creating, and sense reversing
  Sequence* s = new BasicSequence(seq.getName(), content,
                                  seq.getComments(), &_RNY);

  // Send result
  return s;
}

/******************************************************************************/


void SequenceTools::getCDS(Sequence& sequence, bool checkInit, bool checkStop, bool includeInit, bool includeStop)
{
  const CodonAlphabet* alphabet = dynamic_cast<const CodonAlphabet*>(sequence.getAlphabet());
  if (!alphabet)
    throw AlphabetException("SequenceTools::getCDS. Sequence is not a codon sequence.");
  if (checkInit)
  {
    unsigned int i;
    for (i = 0; i < sequence.size() && !alphabet->isInit(sequence[i]); ++i)
    {}
    for (unsigned int j = 0; includeInit ? j < i : j <= i; ++j)
    {
      sequence.deleteElement(j);
    }
  }
  if (checkStop)
  {
    unsigned int i;
    for (i = 0; i < sequence.size() && !alphabet->isStop(sequence[i]); ++i)
    {}
    for (unsigned int j = includeStop ? i + 1 : i; j < sequence.size(); ++j)
    {
      sequence.deleteElement(j);
    }
  }
}

/******************************************************************************/

size_t SequenceTools::findFirstOf(const Sequence& seq, const Sequence& motif, bool strict)
{
  if (motif.size() > seq.size())
    return seq.size();
  for (size_t seqi = 0; seqi < seq.size() - motif.size() + 1; seqi++)
  {
    bool match = false;
    for (size_t moti = 0; moti < motif.size(); moti++)
    {
      if (strict)
      {
        match = seq.getValue(seqi + moti) == motif.getValue(moti);
      }
      else
      {
        match = AlphabetTools::match(seq.getAlphabet(), seq.getValue(seqi + moti), motif.getValue(moti));
      }
      if (!match)
      {
        break;
      }
    }
    if (match)
    {
      return seqi;
    }
  }
  return seq.size();
}

/******************************************************************************/