[jalview.git] / src / jalview / datamodel / AlignedCodonFrame.java

/*
 * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8.2)
 * Copyright (C) 2014 The Jalview Authors
 * 
 * This file is part of Jalview.
 * 
 * Jalview 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.
 *  
 * Jalview 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 Jalview.  If not, see <http://www.gnu.org/licenses/>.
 * The Jalview Authors are detailed in the 'AUTHORS' file.
 */
package jalview.datamodel;

import jalview.util.MapList;

import java.util.ArrayList;
import java.util.List;

/**
 * Stores mapping between the columns of a protein alignment and a DNA alignment
 * and a list of individual codon to amino acid mappings between sequences.
 */
public class AlignedCodonFrame
{

  /**
   * <pre>
   * Aligned nucleotide positions for codons mapped to column positions of of aligned
   * proteins. e.g.
   * codons[3] = [12, 14, 15] means:
   *     column 4 in the protein alignment translates cols 13, 15, 16 in cDNA
   * codons[5] = null means column 6 in the protein alignment is a gap
   * </pre>
   */
  public int[][] codons = null;

  /**
   * Width of protein sequence alignment (implicit assertion that codons.length
   * >= aaWidth)
   */
  public int aaWidth = 0;

  /*
   * TODO: not an ideal solution - we reference the aligned amino acid sequences
   * in order to make insertions on them Better would be dnaAlignment and
   * aaAlignment reference....
   */
  private List<SequenceI> a_aaSeqs = new ArrayList<SequenceI>();

  /*
   * tied array of na Sequence objects.
   */
  private SequenceI[] dnaSeqs = null;

  /*
   * tied array of Mappings to protein sequence Objects and SequenceI[]
   * aaSeqs=null; MapLists where eac maps from the corresponding dnaSeqs element
   * to corresponding aaSeqs element
   */
  private Mapping[] dnaToProt = null;

  /**
   * initialise codon frame with a nominal alignment width
   * 
   * @param aWidth
   */
  public AlignedCodonFrame()
  {
  }

  /**
   * Construct a 'near copy' of the given AlignedCodonFrame, that references the
   * same dataset sequences, but the given protein aligned sequences.
   * 
   * @param acf
   * @param alignment
   * @throws IllegalStateException
   *           if the copied mapping references any dataset not in the alignment
   */
  public AlignedCodonFrame(AlignedCodonFrame acf, SequenceI[] alignment)
  {
    this.codons = acf.codons;
    this.dnaSeqs = acf.dnaSeqs;
    this.dnaToProt = acf.dnaToProt;

    for (SequenceI seq : acf.a_aaSeqs)
    {
      boolean found = false;
      // TODO may not correctly handle the case where the same sequence appears
      // twice in the source alignment i.e. same dataset sequence
      // the copy will reference the first aligned sequence for both
      // ?not solvable if realignment may reorder the sequences
      // or check on sequence name as well????
      for (SequenceI newseq : alignment)
      {
        if (seq.getDatasetSequence() == newseq.getDatasetSequence())
        {
          this.a_aaSeqs.add(newseq);
          found = true;
          break;
        }
      }
      if (!found)
      {
        throw new IllegalStateException("Copying codon mapping for"
                + seq.getSequenceAsString());
      }
    }
  }

  /**
   * ensure that codons array is at least as wide as aslen residues
   * 
   * @param aslen
   * @return (possibly newly expanded) codon array
   */
  public int[][] checkCodonFrameWidth(int aslen)
  {
    // TODO why not codons.length < aslen ?
    // should codons expand if length is 2 or 3 and aslen==2 ?
    if (codons.length <= aslen + 1)
    {
      // probably never have to do this ?
      int[][] c = new int[codons.length + 10][];
      for (int i = 0; i < codons.length; i++)
      {
        c[i] = codons[i];
        codons[i] = null;
      }
      codons = c;
    }
    return codons;
  }

  /**
   * @return width of aligned translated amino acid residues
   */
  public int getaaWidth()
  {
    return aaWidth;
  }

  /**
   * increase aaWidth by one and insert a new aligned codon position space at
   * aspos.
   * 
   * @param aspos
   */
  public void insertAAGap(int aspos, char gapCharacter)
  {
    // this aa appears before the aligned codons at aspos - so shift them in
    // each pair of mapped sequences
    aaWidth++;
    // we actually have to modify the aligned sequences here, so use the
    // a_aaSeqs vector
    for (SequenceI seq : a_aaSeqs)
    {
      seq.insertCharAt(aspos, gapCharacter);
    }

    if (aspos < aaWidth)
    {
      aaWidth++;
      System.arraycopy(codons, aspos, codons, aspos + 1, codons.length
              - aspos - 1);
      codons[aspos] = null; // clear so new codon position can be marked.
    }
  }

  public void setAaWidth(int aapos)
  {
    aaWidth = aapos;
  }

  /**
   * add a mapping between the dataset sequences for the associated dna and
   * protein sequence objects
   * 
   * @param dnaseq
   * @param aaseq
   * @param map
   */
  public void addMap(SequenceI dnaseq, SequenceI aaseq, MapList map)
  {
    int nlen = 1;
    if (dnaSeqs != null)
    {
      nlen = dnaSeqs.length + 1;
    }
    SequenceI[] ndna = new SequenceI[nlen];
    Mapping[] ndtp = new Mapping[nlen];
    if (dnaSeqs != null)
    {
      System.arraycopy(dnaSeqs, 0, ndna, 0, dnaSeqs.length);
      System.arraycopy(dnaToProt, 0, ndtp, 0, dnaSeqs.length);
    }
    dnaSeqs = ndna;
    dnaToProt = ndtp;
    nlen--;
    dnaSeqs[nlen] = (dnaseq.getDatasetSequence() == null) ? dnaseq : dnaseq
            .getDatasetSequence();
    Mapping mp = new Mapping(map);
    // JBPNote DEBUG! THIS !
    // dnaseq.transferAnnotation(aaseq, mp);
    // aaseq.transferAnnotation(dnaseq, new Mapping(map.getInverse()));
    mp.to = (aaseq.getDatasetSequence() == null) ? aaseq : aaseq
            .getDatasetSequence();
    a_aaSeqs.add(aaseq);
    dnaToProt[nlen] = mp;
  }

  public SequenceI[] getdnaSeqs()
  {
    return dnaSeqs;
  }

  public SequenceI[] getAaSeqs()
  {
    if (dnaToProt == null)
    {
      return null;
    }
    SequenceI[] sqs = new SequenceI[dnaToProt.length];
    for (int sz = 0; sz < dnaToProt.length; sz++)
    {
      sqs[sz] = dnaToProt[sz].to;
    }
    return sqs;
  }

  public MapList[] getdnaToProt()
  {
    if (dnaToProt == null)
    {
      return null;
    }
    MapList[] sqs = new MapList[dnaToProt.length];
    for (int sz = 0; sz < dnaToProt.length; sz++)
    {
      sqs[sz] = dnaToProt[sz].map;
    }
    return sqs;
  }

  public Mapping[] getProtMappings()
  {
    return dnaToProt;
  }

  /**
   * 
   * @param sequenceRef
   * @return null or corresponding aaSeq dataset sequence for dnaSeq entry
   */
  public SequenceI getAaForDnaSeq(SequenceI dnaSeqRef)
  {
    return getAaForDnaSeq(dnaSeqRef, true);
  }

  /**
   * Return the corresponding aligned or dataset aa sequence for given dna
   * sequence, null if not found.
   * 
   * @param sequenceRef
   * @param returnDataset
   *          if true, return the aa dataset, else the aligned sequence
   * @return
   */
  public SequenceI getAaForDnaSeq(SequenceI dnaSeqRef, boolean returnDataset)
  {
    if (dnaSeqs == null)
    {
      return null;
    }
    SequenceI dnads = dnaSeqRef.getDatasetSequence();
    for (int ds = 0; ds < dnaSeqs.length; ds++)
    {
      if (dnaSeqs[ds] == dnaSeqRef || dnaSeqs[ds] == dnads)
      {
        if (returnDataset)
        {
          return dnaToProt[ds].to;
        }
        else
        {
          // TODO very fragile - depends on dnaSeqs, dnaToProt, a_aaSeqs moving
          // in parallel; revise data model to guarantee this
          return a_aaSeqs.get(ds);
        }
      }
    }
    return null;
  }

  /**
   * 
   * @param sequenceRef
   * @return null or corresponding aaSeq entry for dnaSeq entry
   */
  public SequenceI getDnaForAaSeq(SequenceI aaSeqRef)
  {
    if (dnaToProt == null)
    {
      return null;
    }
    SequenceI aads = aaSeqRef.getDatasetSequence();
    for (int as = 0; as < dnaToProt.length; as++)
    {
      if (dnaToProt[as].to == aaSeqRef || dnaToProt[as].to == aads)
      {
        return dnaSeqs[as];
      }
    }
    return null;
  }

  /**
   * test to see if codon frame involves seq in any way
   * 
   * @param seq
   *          a nucleotide or protein sequence
   * @return true if a mapping exists to or from this sequence to any translated
   *         sequence
   */
  public boolean involvesSequence(SequenceI seq)
  {
    return getAaForDnaSeq(seq) != null || getDnaForAaSeq(seq) != null;
  }

  /**
   * Add search results for regions in other sequences that translate or are
   * translated from a particular position in seq
   * 
   * @param seq
   * @param index
   *          position in seq
   * @param results
   *          where highlighted regions go
   */
  public void markMappedRegion(SequenceI seq, int index,
          SearchResults results)
  {
    if (dnaToProt == null)
    {
      return;
    }
    int[] codon;
    SequenceI ds = seq.getDatasetSequence();
    for (int mi = 0; mi < dnaToProt.length; mi++)
    {
      if (dnaSeqs[mi] == seq || dnaSeqs[mi] == ds)
      {
        // DEBUG System.err.println("dna pos "+index);
        codon = dnaToProt[mi].map.locateInTo(index, index);
        if (codon != null)
        {
          for (int i = 0; i < codon.length; i += 2)
          {
            results.addResult(dnaToProt[mi].to, codon[i], codon[i + 1]);
          }
        }
      }
      else if (dnaToProt[mi].to == seq || dnaToProt[mi].to == ds)
      {
        // DEBUG System.err.println("aa pos "+index);
        {
          codon = dnaToProt[mi].map.locateInFrom(index, index);
          if (codon != null)
          {
            for (int i = 0; i < codon.length; i += 2)
            {
              results.addResult(dnaSeqs[mi], codon[i], codon[i + 1]);
            }
          }
        }
      }
    }
  }

  /**
   * Returns the DNA codon positions (base 1) for the given position (base 1) in
   * a mapped protein sequence, or null if no mapping is found.
   * 
   * Intended for use in aligning cDNA to match aligned protein. Only the first
   * mapping found is returned, so not suitable for use if multiple protein
   * sequences are mapped to the same cDNA (but aligning cDNA as protein is
   * ill-defined for this case anyway).
   * 
   * @param seq
   *          the DNA dataset sequence
   * @param aaPos
   *          residue position (base 1) in a protein sequence
   * @return
   */
  public int[] getDnaPosition(SequenceI seq, int aaPos)
  {
    /*
     * Adapted from markMappedRegion().
     */
    MapList ml = null;
    for (int i = 0; i < dnaToProt.length; i++)
    {
      if (dnaSeqs[i] == seq)
      {
        ml = getdnaToProt()[i];
        break;
      }
    }
    return ml == null ? null : ml.locateInFrom(aaPos, aaPos);
  }
}