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;
29 import java.awt.Color;
30 import java.util.Hashtable;
31 import java.util.List;
33 import java.util.Vector;
36 * Calculates conservation values for a given set of sequences
41 public class Conservation
43 SequenceI[] sequences;
49 Vector<int[]> seqNums; // vector of int vectors where first is sequence
52 int maxLength = 0; // used by quality calcs
54 boolean seqNumsChanged = false; // updated after any change via calcSeqNum;
56 Map<String, Integer>[] total;
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 private Vector<Double> quality;
67 /** Stores maximum and minimum values of quality values */
68 private double[] qualityRange = new double[2];
70 private Sequence consSequence;
72 private int threshold;
74 private String name = "";
76 private int[][] cons2;
78 private String[] consSymbs;
81 * Creates a new Conservation object.
84 * Name of conservation
86 * to count the residues in residueHash(). commonly used value is 3
88 * sequences to be used in calculation
90 * start residue position
92 * end residue position
94 public Conservation(String name, int threshold,
95 List<SequenceI> sequences, int start, int end)
98 this.threshold = threshold;
102 maxLength = end - start + 1; // default width includes bounds of
105 int s, sSize = sequences.size();
106 SequenceI[] sarray = new SequenceI[sSize];
107 this.sequences = sarray;
110 for (s = 0; s < sSize; s++)
112 sarray[s] = sequences.get(s);
113 if (sarray[s].getLength() > maxLength)
115 maxLength = sarray[s].getLength();
118 } catch (ArrayIndexOutOfBoundsException ex)
120 // bail - another thread has modified the sequence array, so the current
121 // calculation is probably invalid.
122 this.sequences = new SequenceI[0];
128 * Translate sequence i into a numerical representation and store it in the
129 * i'th position of the seqNums array.
133 private void calcSeqNum(int i)
135 String sq = null; // for dumb jbuilder not-inited exception warning
138 int sSize = sequences.length;
140 if ((i > -1) && (i < sSize))
142 sq = sequences[i].getSequenceAsString();
144 if (seqNums.size() <= i)
146 seqNums.addElement(new int[sq.length() + 1]);
149 if (sq.hashCode() != seqNums.elementAt(i)[0])
153 seqNumsChanged = true;
161 sqnum = new int[len + 1]; // better to always make a new array -
162 // sequence can change its length
163 sqnum[0] = sq.hashCode();
165 for (j = 1; j <= len; j++)
167 sqnum[j] = jalview.schemes.ResidueProperties.aaIndex[sq
171 seqNums.setElementAt(sqnum, i);
175 System.out.println("SEQUENCE HAS BEEN DELETED!!!");
180 // JBPNote INFO level debug
182 .println("ERROR: calcSeqNum called with out of range sequence index for Alignment\n");
187 * Calculates the conservation values for given set of sequences
189 public void calculate()
191 int thresh, j, jSize = sequences.length;
192 int[] values; // Replaces residueHash
195 total = new Hashtable[maxLength];
197 for (int i = start; i <= end; i++)
199 values = new int[255];
201 for (j = 0; j < jSize; j++)
203 if (sequences[j].getLength() > i)
205 c = sequences[j].getCharAt(i);
208 { // lookup the base aa code symbol
209 c = (char) ResidueProperties.aaIndex[sequences[j].getCharAt(i)];
216 // recover canonical aa symbol
217 c = ResidueProperties.aa[c].charAt(0);
222 // original behaviour - operate on ascii symbols directly
223 // No need to check if its a '-'
224 if (c == '.' || c == ' ')
239 // What is the count threshold to count the residues in residueHash()
240 thresh = (threshold * (jSize)) / 100;
242 // loop over all the found residues
243 Hashtable<String, Integer> resultHash = new Hashtable<String, Integer>();
244 for (char v = '-'; v < 'Z'; v++)
247 if (values[v] > thresh)
249 String res = String.valueOf(v);
251 // Now loop over the properties
252 for (String type : ResidueProperties.propHash.keySet())
254 Map<String, Integer> ht = ResidueProperties.propHash.get(type);
256 // Have we ticked this before?
257 if (!resultHash.containsKey(type))
259 if (ht.containsKey(res))
261 resultHash.put(type, ht.get(res));
265 resultHash.put(type, ht.get("-"));
268 else if (!resultHash.get(type).equals(ht.get(res)))
270 resultHash.put(type, new Integer(-1));
276 if (total.length > 0)
278 total[i - start] = resultHash;
283 /*****************************************************************************
284 * count conservation for the j'th column of the alignment
286 * @return { gap count, conserved residue count}
288 public int[] countConsNGaps(int j)
293 int[] r = new int[2];
295 int i, iSize = sequences.length;
298 for (i = 0; i < iSize; i++)
300 if (j >= sequences[i].getLength())
306 c = sequences[i].getCharAt(j); // gaps do not have upper/lower case
308 if (jalview.util.Comparison.isGap((c)))
329 r[0] = (nres == cons) ? 1 : 0;
336 * Returns the upper-cased character if between 'a' and 'z', else the
342 char toUpperCase(char c)
344 if ('a' <= c && c <= 'z')
346 c -= (32); // 32 = 'a' - 'A'
352 * Calculates the conservation sequence
355 * if true, positive conservation; false calculates negative
357 * @param percentageGaps
358 * commonly used value is 25
360 public void verdict(boolean consflag, float percentageGaps)
362 StringBuffer consString = new StringBuffer();
364 // NOTE THIS SHOULD CHECK IF THE CONSEQUENCE ALREADY
365 // EXISTS AND NOT OVERWRITE WITH '-', BUT THIS CASE
366 // DOES NOT EXIST IN JALVIEW 2.1.2
367 for (int i = 0; i < start; i++)
369 consString.append('-');
371 consSymbs = new String[end - start + 1];
372 for (int i = start; i <= end; i++)
374 int[] gapcons = countConsNGaps(i);
375 int totGaps = gapcons[1];
376 float pgaps = ((float) totGaps * 100) / sequences.length;
377 consSymbs[i - start] = new String();
379 if (percentageGaps > pgaps)
381 Map<String, Integer> resultHash = total[i - start];
382 // Now find the verdict
384 for (String type : resultHash.keySet())
386 int result = resultHash.get(type).intValue();
387 // Do we want to count +ve conservation or +ve and -ve cons.?
392 consSymbs[i - start] = type + " " + consSymbs[i - start];
402 consSymbs[i - start] = consSymbs[i - start] + " !" + type;
406 consSymbs[i - start] = type + " " + consSymbs[i - start];
415 consString.append(count); // Conserved props!=Identity
419 consString.append((gapcons[0] == 1) ? "*" : "+");
424 consString.append('-');
428 consSequence = new Sequence(name, consString.toString(), start, end);
434 * @return Conservation sequence
436 public Sequence getConsSequence()
441 // From Alignment.java in jalview118
442 public void findQuality()
444 findQuality(0, maxLength - 1);
450 private void percentIdentity2()
452 seqNums = new Vector<int[]>();
454 int i = 0, iSize = sequences.length;
455 // Do we need to calculate this again?
456 for (i = 0; i < iSize; i++)
461 if ((cons2 == null) || seqNumsChanged)
463 cons2 = new int[maxLength][24];
465 // Initialize the array
466 for (int j = 0; j < 24; j++)
468 for (i = 0; i < maxLength; i++)
477 while (j < sequences.length)
479 sqnum = seqNums.elementAt(j);
481 for (i = 1; i < sqnum.length; i++)
483 cons2[i - 1][sqnum[i]]++;
486 for (i = sqnum.length - 1; i < maxLength; i++)
488 cons2[i][23]++; // gap count
497 * for (int i=start; i <= end; i++) { int max = -1000; int maxi = -1; int
500 * for (int j=0;j<24;j++) { if (cons2[i][j] > max) { max = cons2[i][j];
501 * maxi = i; maxj = j; } } }
507 * Calculates the quality of the set of sequences
514 public void findQuality(int startRes, int endRes)
516 quality = new Vector<Double>();
519 int[][] BLOSUM62 = ResidueProperties.getBLOSUM62();
521 // Loop over columns // JBPNote Profiling info
522 // long ts = System.currentTimeMillis();
523 // long te = System.currentTimeMillis();
526 int size = seqNums.size();
527 int[] lengths = new int[size];
528 double tot, bigtot, sr, tmp;
530 int l, j, i, ii, i2, k, seqNum;
532 for (l = 0; l < size; l++)
534 lengths[l] = seqNums.elementAt(l).length - 1;
537 for (j = startRes; j <= endRes; j++)
541 // First Xr = depends on column only
544 for (ii = 0; ii < 24; ii++)
548 for (i2 = 0; i2 < 24; i2++)
550 x[ii] += (((double) cons2[j][i2] * BLOSUM62[ii][i2]) + 4);
556 // Now calculate D for each position and sum
557 for (k = 0; k < size; k++)
561 seqNum = (j < lengths[k]) ? seqNums.elementAt(k)[j + 1]
562 : 23; // Sequence, or gap at the end
564 // This is a loop over r
565 for (i = 0; i < 23; i++)
569 sr = (double) BLOSUM62[i][seqNum] + 4;
571 // Calculate X with another loop over residues
572 // System.out.println("Xi " + i + " " + x[i] + " " + sr);
575 tot += (xx[i] * xx[i]);
578 bigtot += Math.sqrt(tot);
581 // This is the quality for one column
587 // bigtot = bigtot * (size-cons2[j][23])/size;
588 quality.addElement(new Double(bigtot));
590 // Need to normalize by gaps
593 double newmax = -10000;
595 for (j = startRes; j <= endRes; j++)
597 tmp = quality.elementAt(j).doubleValue();
598 tmp = ((max - tmp) * (size - cons2[j][23])) / size;
600 // System.out.println(tmp+ " " + j);
601 quality.setElementAt(new Double(tmp), j);
609 // System.out.println("Quality " + s);
610 qualityRange[0] = 0D;
611 qualityRange[1] = newmax;
615 * Complete the given consensus and quuality annotation rows. Note: currently
616 * this method will enlarge the given annotation row if it is too small,
617 * otherwise will leave its length unchanged.
619 * @param conservation
620 * conservation annotation row
622 * (optional - may be null)
624 * first column for conservation
626 * extent of conservation
628 public void completeAnnotations(AlignmentAnnotation conservation,
629 AlignmentAnnotation quality2, int istart, int alWidth)
631 char[] sequence = getConsSequence().getSequence();
643 maxB = 0f - minB; // scalable range for colouring both Conservation and
653 if (conservation != null && conservation.annotations != null
654 && conservation.annotations.length < alWidth)
656 conservation.annotations = new Annotation[alWidth];
659 if (quality2 != null)
661 quality2.graphMax = (float) qualityRange[1];
662 if (quality2.annotations != null
663 && quality2.annotations.length < alWidth)
665 quality2.annotations = new Annotation[alWidth];
667 qmin = (float) qualityRange[0];
668 qmax = (float) qualityRange[1];
671 for (int i = istart; i < alWidth; i++)
677 if (Character.isDigit(c))
690 if (conservation != null)
692 float vprop = value - min;
694 int consp = i - start;
695 String conssym = (value > 0 && consp > -1 && consp < consSymbs.length) ? consSymbs[consp]
697 conservation.annotations[i] = new Annotation(String.valueOf(c),
698 conssym, ' ', value, new Color(minR + (maxR * vprop), minG
699 + (maxG * vprop), minB + (maxB * vprop)));
703 if (quality2 != null)
705 value = quality.elementAt(i).floatValue();
706 float vprop = value - qmin;
708 quality2.annotations[i] = new Annotation(" ",
709 String.valueOf(value), ' ', value, new Color(minR
710 + (maxR * vprop), minG + (maxG * vprop), minB
717 * construct and call the calculation methods on a new Conservation object
720 * - name of conservation
722 * - minimum number of conserved residues needed to indicate
723 * conservation (typically 3)
726 * first column in calculation window
728 * last column in calculation window
730 * positive (true) or negative (false) conservation
731 * @param consPercGaps
732 * percentage of gaps tolerated in column
734 * flag indicating if alignment quality should be calculated
735 * @return Conservation object ready for use in visualization
737 public static Conservation calculateConservation(String name,
738 int threshold, List<SequenceI> seqs, int start, int end,
739 boolean posOrNeg, int consPercGaps, boolean calcQuality)
741 Conservation cons = new Conservation(name, threshold, seqs, start, end);
743 cons.verdict(posOrNeg, consPercGaps);