2 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3 * Copyright (C) $$Year-Rel$$ The Jalview Authors
5 * This file is part of Jalview.
7 * Jalview is free software: you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation, either version 3
10 * of the License, or (at your option) any later version.
12 * Jalview is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15 * PURPOSE. See the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19 * The Jalview Authors are detailed in the 'AUTHORS' file.
21 package jalview.analysis;
23 import jalview.datamodel.AlignmentAnnotation;
24 import jalview.datamodel.Annotation;
25 import jalview.datamodel.Sequence;
26 import jalview.datamodel.SequenceI;
27 import jalview.schemes.ResidueProperties;
28 import jalview.util.Comparison;
30 import java.awt.Color;
31 import java.util.Enumeration;
32 import java.util.Hashtable;
33 import java.util.List;
34 import java.util.Vector;
37 * Calculates conservation values for a given set of sequences
42 public class Conservation
44 SequenceI[] sequences;
50 Vector seqNums; // vector of int vectors where first is sequence checksum
52 int maxLength = 0; // used by quality calcs
54 boolean seqNumsChanged = false; // updated after any change via calcSeqNum;
58 boolean canonicaliseAa = true; // if true then conservation calculation will
60 // map all symbols to canonical aa numbering
61 // rather than consider conservation of that
64 /** Stores calculated quality values */
65 public Vector quality;
67 /** Stores maximum and minimum values of quality values */
68 public Double[] qualityRange = new Double[2];
70 String consString = "";
72 Sequence consSequence;
82 private String[] consSymbs;
85 * Creates a new Conservation object.
88 * Name of conservation
90 * hash of properties for each symbol
92 * to count the residues in residueHash(). commonly used value is 3
94 * sequences to be used in calculation
96 * start residue position
98 * end residue position
100 public Conservation(String name, Hashtable propHash, int threshold,
101 List<SequenceI> sequences, int start, int end)
104 this.propHash = propHash;
105 this.threshold = threshold;
109 maxLength = end - start + 1; // default width includes bounds of
112 int s, sSize = sequences.size();
113 SequenceI[] sarray = new SequenceI[sSize];
114 this.sequences = sarray;
117 for (s = 0; s < sSize; s++)
119 sarray[s] = sequences.get(s);
120 if (sarray[s].getLength() > maxLength)
122 maxLength = sarray[s].getLength();
125 } catch (ArrayIndexOutOfBoundsException ex)
127 // bail - another thread has modified the sequence array, so the current
128 // calculation is probably invalid.
129 this.sequences = new SequenceI[0];
135 * Translate sequence i into a numerical representation and store it in the
136 * i'th position of the seqNums array.
140 private void calcSeqNum(int i)
142 String sq = null; // for dumb jbuilder not-inited exception warning
145 int sSize = sequences.length;
147 if ((i > -1) && (i < sSize))
149 sq = sequences[i].getSequenceAsString();
151 if (seqNums.size() <= i)
153 seqNums.addElement(new int[sq.length() + 1]);
156 if (sq.hashCode() != ((int[]) seqNums.elementAt(i))[0])
160 seqNumsChanged = true;
168 sqnum = new int[len + 1]; // better to always make a new array -
169 // sequence can change its length
170 sqnum[0] = sq.hashCode();
172 for (j = 1; j <= len; j++)
174 sqnum[j] = ResidueProperties.aaIndex[sq
178 seqNums.setElementAt(sqnum, i);
182 System.out.println("SEQUENCE HAS BEEN DELETED!!!");
187 // JBPNote INFO level debug
189 .println("ERROR: calcSeqNum called with out of range sequence index for Alignment\n");
194 * Calculates the conservation values for given set of sequences
196 public void calculate()
198 Hashtable resultHash, ht;
199 int thresh, j, jSize = sequences.length;
200 int[] values; // Replaces residueHash
201 String type, res = null;
203 Enumeration enumeration2;
205 total = new Hashtable[maxLength];
207 for (int i = start; i <= end; i++)
209 values = new int[255];
211 for (j = 0; j < jSize; j++)
213 if (sequences[j].getLength() > i)
215 c = sequences[j].getCharAt(i);
218 { // lookup the base aa code symbol
219 c = (char) ResidueProperties.aaIndex[sequences[j]
227 // recover canonical aa symbol
228 c = ResidueProperties.aa[c].charAt(0);
233 // original behaviour - operate on ascii symbols directly
234 // No need to check if its a '-'
235 if (c == '.' || c == ' ')
240 if (!canonicaliseAa && 'a' <= c && c <= 'z')
242 c -= (32); // 32 = 'a' - 'A'
253 // What is the count threshold to count the residues in residueHash()
254 thresh = (threshold * (jSize)) / 100;
256 // loop over all the found residues
257 resultHash = new Hashtable();
258 for (char v = '-'; v < 'Z'; v++)
261 if (values[v] > thresh)
263 res = String.valueOf(v);
265 // Now loop over the properties
266 enumeration2 = propHash.keys();
268 while (enumeration2.hasMoreElements())
270 type = (String) enumeration2.nextElement();
271 ht = (Hashtable) propHash.get(type);
273 // Have we ticked this before?
274 if (!resultHash.containsKey(type))
276 if (ht.containsKey(res))
278 resultHash.put(type, ht.get(res));
282 resultHash.put(type, ht.get("-"));
285 else if (((Integer) resultHash.get(type)).equals(ht
288 resultHash.put(type, new Integer(-1));
294 if (total.length > 0)
296 total[i - start] = resultHash;
301 /*****************************************************************************
302 * count conservation for the j'th column of the alignment
304 * @return { gap count, conserved residue count}
306 public int[] countConsNGaps(int j)
311 int[] r = new int[2];
313 int i, iSize = sequences.length;
316 for (i = 0; i < iSize; i++)
318 if (j >= sequences[i].getLength())
324 c = sequences[i].getCharAt(j); // gaps do not have upper/lower case
326 if (Comparison.isGap((c)))
346 r[0] = (nres == cons) ? 1 : 0;
353 * Calculates the conservation sequence
356 * if true, poitiveve conservation; false calculates negative
358 * @param percentageGaps
359 * commonly used value is 25
361 public void verdict(boolean consflag, float percentageGaps)
363 StringBuffer consString = new StringBuffer();
369 Hashtable resultHash;
370 Enumeration enumeration;
372 // NOTE THIS SHOULD CHECK IF THE CONSEQUENCE ALREADY
373 // EXISTS AND NOT OVERWRITE WITH '-', BUT THIS CASE
374 // DOES NOT EXIST IN JALVIEW 2.1.2
375 for (int i = 0; i < start; i++)
377 consString.append('-');
379 consSymbs = new String[end-start+1];
380 for (int i = start; i <= end; i++)
382 gapcons = countConsNGaps(i);
383 totGaps = gapcons[1];
384 pgaps = ((float) totGaps * 100) / sequences.length;
385 consSymbs[i-start]=new String();
387 if (percentageGaps > pgaps)
389 resultHash = total[i - start];
390 // Now find the verdict
392 enumeration = resultHash.keys();
394 while (enumeration.hasMoreElements())
396 type = (String) enumeration.nextElement();
397 result = (Integer) resultHash.get(type);
398 // Do we want to count +ve conservation or +ve and -ve cons.?
401 if (result.intValue() == 1)
403 consSymbs[i-start] = type+" "+consSymbs[i-start];
409 if (result.intValue() != -1)
412 if (result.intValue()==0) {
413 consSymbs[i-start] = consSymbs[i-start]+ " !"+type;
415 consSymbs[i-start] = type+" "+consSymbs[i-start];
426 consString.append(count); // Conserved props!=Identity
430 consString.append((gapcons[0] == 1) ? "*" : "+");
435 consString.append('-');
439 consSequence = new Sequence(name, consString.toString(), start, end);
445 * @return Conservation sequence
447 public Sequence getConsSequence()
452 // From Alignment.java in jalview118
453 public void findQuality()
455 findQuality(0, maxLength - 1);
461 private void percentIdentity2()
463 seqNums = new Vector();
465 int i = 0, iSize = sequences.length;
466 // Do we need to calculate this again?
467 for (i = 0; i < iSize; i++)
472 if ((cons2 == null) || seqNumsChanged)
474 cons2 = new int[maxLength][24];
476 // Initialize the array
477 for (int j = 0; j < 24; j++)
479 for (i = 0; i < maxLength; i++)
488 while (j < sequences.length)
490 sqnum = (int[]) seqNums.elementAt(j);
492 for (i = 1; i < sqnum.length; i++)
494 cons2[i - 1][sqnum[i]]++;
497 for (i = sqnum.length - 1; i < maxLength; i++)
499 cons2[i][23]++; // gap count
508 * for (int i=start; i <= end; i++) { int max = -1000; int maxi = -1; int
511 * for (int j=0;j<24;j++) { if (cons2[i][j] > max) { max = cons2[i][j];
512 * maxi = i; maxj = j; } } }
518 * Calculates the quality of the set of sequences
525 public void findQuality(int start, int end)
527 quality = new Vector();
530 int[][] BLOSUM62 = ResidueProperties.getBLOSUM62();
532 // Loop over columns // JBPNote Profiling info
533 // long ts = System.currentTimeMillis();
534 // long te = System.currentTimeMillis();
537 int size = seqNums.size();
538 int[] lengths = new int[size];
539 double tot, bigtot, sr, tmp;
541 int l, j, i, ii, i2, k, seqNum;
543 for (l = 0; l < size; l++)
545 lengths[l] = ((int[]) seqNums.elementAt(l)).length - 1;
548 for (j = start; j <= end; j++)
552 // First Xr = depends on column only
555 for (ii = 0; ii < 24; ii++)
559 for (i2 = 0; i2 < 24; i2++)
561 x[ii] += (((double) cons2[j][i2] * BLOSUM62[ii][i2]) + 4);
567 // Now calculate D for each position and sum
568 for (k = 0; k < size; k++)
572 seqNum = (j < lengths[k]) ? ((int[]) seqNums.elementAt(k))[j + 1]
573 : 23; // Sequence, or gap at the end
575 // This is a loop over r
576 for (i = 0; i < 23; i++)
580 sr = (double) BLOSUM62[i][seqNum] + 4;
582 // Calculate X with another loop over residues
583 // System.out.println("Xi " + i + " " + x[i] + " " + sr);
586 tot += (xx[i] * xx[i]);
589 bigtot += Math.sqrt(tot);
592 // This is the quality for one column
598 // bigtot = bigtot * (size-cons2[j][23])/size;
599 quality.addElement(new Double(bigtot));
601 // Need to normalize by gaps
604 double newmax = -10000;
606 for (j = start; j <= end; j++)
608 tmp = ((Double) quality.elementAt(j)).doubleValue();
609 tmp = ((max - tmp) * (size - cons2[j][23])) / size;
611 // System.out.println(tmp+ " " + j);
612 quality.setElementAt(new Double(tmp), j);
620 // System.out.println("Quality " + s);
621 qualityRange[0] = new Double(0);
622 qualityRange[1] = new Double(newmax);
626 * complete the given consensus and quuality annotation rows. Note: currently
627 * this method will enlarge the given annotation row if it is too small,
628 * otherwise will leave its length unchanged.
630 * @param conservation
631 * conservation annotation row
633 * (optional - may be null)
635 * first column for conservation
637 * extent of conservation
639 public void completeAnnotations(AlignmentAnnotation conservation,
640 AlignmentAnnotation quality2, int istart, int alWidth)
642 char[] sequence = getConsSequence().getSequence();
654 maxB = 0f - minB; // scalable range for colouring both Conservation and
664 if (conservation.annotations != null
665 && conservation.annotations.length < alWidth)
667 conservation.annotations = new Annotation[alWidth];
670 if (quality2 != null)
672 quality2.graphMax = qualityRange[1].floatValue();
673 if (quality2.annotations != null
674 && quality2.annotations.length < alWidth)
676 quality2.annotations = new Annotation[alWidth];
678 qmin = qualityRange[0].floatValue();
679 qmax = qualityRange[1].floatValue();
682 for (int i = 0; i < alWidth; i++)
688 if (Character.isDigit(c))
701 float vprop = value - min;
703 conservation.annotations[i] = new Annotation(String.valueOf(c),
704 consSymbs[i-start], ' ', value, new Color(minR
705 + (maxR * vprop), minG + (maxG * vprop), minB
709 if (quality2 != null)
711 value = ((Double) quality.elementAt(i)).floatValue();
712 vprop = value - qmin;
714 quality2.annotations[i] = new Annotation(" ",
715 String.valueOf(value), ' ', value, new Color(minR
716 + (maxR * vprop), minG + (maxG * vprop), minB
723 * construct and call the calculation methods on a new Conservation object
726 * - name of conservation
728 * - hash table of properties for each amino acid (normally
729 * ResidueProperties.propHash)
731 * - minimum number of conserved residues needed to indicate
732 * conservation (typically 3)
735 * first column in calculation window
737 * last column in calculation window
739 * positive (true) or negative (false) conservation
740 * @param consPercGaps
741 * percentage of gaps tolerated in column
743 * flag indicating if alignment quality should be calculated
744 * @return Conservation object ready for use in visualization
746 public static Conservation calculateConservation(String name,
747 Hashtable consHash, int threshold, List<SequenceI> seqs,
748 int start, int end, boolean posOrNeg, int consPercGaps,
751 Conservation cons = new Conservation(name, consHash, threshold, seqs,
753 return calculateConservation(cons, posOrNeg, consPercGaps, calcQuality);
758 * positive (true) or negative (false) conservation
759 * @param consPercGaps
760 * percentage of gaps tolerated in column
762 * flag indicating if alignment quality should be calculated
763 * @return Conservation object ready for use in visualization
765 public static Conservation calculateConservation(Conservation cons,
766 boolean b, int consPercGaps, boolean calcQuality)
769 cons.verdict(b, consPercGaps);