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 Hashtable<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 public Vector<Double> quality;
67 /** Stores maximum and minimum values of quality values */
68 private double[] qualityRange = new double[2];
70 Sequence consSequence;
72 Map<String, Map<String, Integer>> propHash;
80 private String[] consSymbs;
83 * Creates a new Conservation object.
86 * Name of conservation
88 * hash of properties for each symbol
90 * to count the residues in residueHash(). commonly used value is 3
92 * sequences to be used in calculation
94 * start residue position
96 * end residue position
98 public Conservation(String name, Hashtable propHash, int threshold,
99 List<SequenceI> sequences, int start, int end)
102 this.propHash = propHash;
103 this.threshold = threshold;
107 maxLength = end - start + 1; // default width includes bounds of
110 int s, sSize = sequences.size();
111 SequenceI[] sarray = new SequenceI[sSize];
112 this.sequences = sarray;
115 for (s = 0; s < sSize; s++)
117 sarray[s] = sequences.get(s);
118 if (sarray[s].getLength() > maxLength)
120 maxLength = sarray[s].getLength();
123 } catch (ArrayIndexOutOfBoundsException ex)
125 // bail - another thread has modified the sequence array, so the current
126 // calculation is probably invalid.
127 this.sequences = new SequenceI[0];
133 * Translate sequence i into a numerical representation and store it in the
134 * i'th position of the seqNums array.
138 private void calcSeqNum(int i)
140 String sq = null; // for dumb jbuilder not-inited exception warning
143 int sSize = sequences.length;
145 if ((i > -1) && (i < sSize))
147 sq = sequences[i].getSequenceAsString();
149 if (seqNums.size() <= i)
151 seqNums.addElement(new int[sq.length() + 1]);
154 if (sq.hashCode() != seqNums.elementAt(i)[0])
158 seqNumsChanged = true;
166 sqnum = new int[len + 1]; // better to always make a new array -
167 // sequence can change its length
168 sqnum[0] = sq.hashCode();
170 for (j = 1; j <= len; j++)
172 sqnum[j] = jalview.schemes.ResidueProperties.aaIndex[sq
176 seqNums.setElementAt(sqnum, i);
180 System.out.println("SEQUENCE HAS BEEN DELETED!!!");
185 // JBPNote INFO level debug
187 .println("ERROR: calcSeqNum called with out of range sequence index for Alignment\n");
192 * Calculates the conservation values for given set of sequences
194 public void calculate()
196 int thresh, j, jSize = sequences.length;
197 int[] values; // Replaces residueHash
200 total = new Hashtable[maxLength];
202 for (int i = start; i <= end; i++)
204 values = new int[255];
206 for (j = 0; j < jSize; j++)
208 if (sequences[j].getLength() > i)
210 c = sequences[j].getCharAt(i);
213 { // lookup the base aa code symbol
214 c = (char) ResidueProperties.aaIndex[sequences[j].getCharAt(i)];
221 // recover canonical aa symbol
222 c = ResidueProperties.aa[c].charAt(0);
227 // original behaviour - operate on ascii symbols directly
228 // No need to check if its a '-'
229 if (c == '.' || c == ' ')
244 // What is the count threshold to count the residues in residueHash()
245 thresh = (threshold * (jSize)) / 100;
247 // loop over all the found residues
248 Hashtable<String, Integer> resultHash = new Hashtable<String, Integer>();
249 for (char v = '-'; v < 'Z'; v++)
252 if (values[v] > thresh)
254 String res = String.valueOf(v);
256 // Now loop over the properties
257 for (String type : propHash.keySet())
259 Hashtable<String, Integer> ht = (Hashtable<String, Integer>) propHash
262 // Have we ticked this before?
263 if (!resultHash.containsKey(type))
265 if (ht.containsKey(res))
267 resultHash.put(type, ht.get(res));
271 resultHash.put(type, ht.get("-"));
274 else if (!resultHash.get(type).equals(ht.get(res)))
276 resultHash.put(type, new Integer(-1));
282 if (total.length > 0)
284 total[i - start] = resultHash;
289 /*****************************************************************************
290 * count conservation for the j'th column of the alignment
292 * @return { gap count, conserved residue count}
294 public int[] countConsNGaps(int j)
299 int[] r = new int[2];
301 int i, iSize = sequences.length;
304 for (i = 0; i < iSize; i++)
306 if (j >= sequences[i].getLength())
312 c = sequences[i].getCharAt(j); // gaps do not have upper/lower case
314 if (jalview.util.Comparison.isGap((c)))
335 r[0] = (nres == cons) ? 1 : 0;
342 * Returns the upper-cased character if between 'a' and 'z', else the
348 char toUpperCase(char c)
350 if ('a' <= c && c <= 'z')
352 c -= (32); // 32 = 'a' - 'A'
358 * Calculates the conservation sequence
361 * if true, positive conservation; false calculates negative
363 * @param percentageGaps
364 * commonly used value is 25
366 public void verdict(boolean consflag, float percentageGaps)
368 StringBuffer consString = new StringBuffer();
370 // NOTE THIS SHOULD CHECK IF THE CONSEQUENCE ALREADY
371 // EXISTS AND NOT OVERWRITE WITH '-', BUT THIS CASE
372 // DOES NOT EXIST IN JALVIEW 2.1.2
373 for (int i = 0; i < start; i++)
375 consString.append('-');
377 consSymbs = new String[end - start + 1];
378 for (int i = start; i <= end; i++)
380 int[] gapcons = countConsNGaps(i);
381 int totGaps = gapcons[1];
382 float pgaps = ((float) totGaps * 100) / sequences.length;
383 consSymbs[i - start] = new String();
385 if (percentageGaps > pgaps)
387 Hashtable<String, Integer> resultHash = total[i - start];
388 // Now find the verdict
390 for (String type : resultHash.keySet())
392 int result = resultHash.get(type).intValue();
393 // Do we want to count +ve conservation or +ve and -ve cons.?
398 consSymbs[i - start] = type + " " + consSymbs[i - start];
408 consSymbs[i - start] = consSymbs[i - start] + " !" + type;
412 consSymbs[i - start] = type + " " + consSymbs[i - start];
421 consString.append(count); // Conserved props!=Identity
425 consString.append((gapcons[0] == 1) ? "*" : "+");
430 consString.append('-');
434 consSequence = new Sequence(name, consString.toString(), start, end);
440 * @return Conservation sequence
442 public Sequence getConsSequence()
447 // From Alignment.java in jalview118
448 public void findQuality()
450 findQuality(0, maxLength - 1);
456 private void percentIdentity2()
458 seqNums = new Vector<int[]>();
460 int i = 0, iSize = sequences.length;
461 // Do we need to calculate this again?
462 for (i = 0; i < iSize; i++)
467 if ((cons2 == null) || seqNumsChanged)
469 cons2 = new int[maxLength][24];
471 // Initialize the array
472 for (int j = 0; j < 24; j++)
474 for (i = 0; i < maxLength; i++)
483 while (j < sequences.length)
485 sqnum = seqNums.elementAt(j);
487 for (i = 1; i < sqnum.length; i++)
489 cons2[i - 1][sqnum[i]]++;
492 for (i = sqnum.length - 1; i < maxLength; i++)
494 cons2[i][23]++; // gap count
503 * for (int i=start; i <= end; i++) { int max = -1000; int maxi = -1; int
506 * for (int j=0;j<24;j++) { if (cons2[i][j] > max) { max = cons2[i][j];
507 * maxi = i; maxj = j; } } }
513 * Calculates the quality of the set of sequences
520 public void findQuality(int startRes, int endRes)
522 quality = new Vector<Double>();
525 int[][] BLOSUM62 = ResidueProperties.getBLOSUM62();
527 // Loop over columns // JBPNote Profiling info
528 // long ts = System.currentTimeMillis();
529 // long te = System.currentTimeMillis();
532 int size = seqNums.size();
533 int[] lengths = new int[size];
534 double tot, bigtot, sr, tmp;
536 int l, j, i, ii, i2, k, seqNum;
538 for (l = 0; l < size; l++)
540 lengths[l] = seqNums.elementAt(l).length - 1;
543 for (j = startRes; j <= endRes; j++)
547 // First Xr = depends on column only
550 for (ii = 0; ii < 24; ii++)
554 for (i2 = 0; i2 < 24; i2++)
556 x[ii] += (((double) cons2[j][i2] * BLOSUM62[ii][i2]) + 4);
562 // Now calculate D for each position and sum
563 for (k = 0; k < size; k++)
567 seqNum = (j < lengths[k]) ? seqNums.elementAt(k)[j + 1]
568 : 23; // Sequence, or gap at the end
570 // This is a loop over r
571 for (i = 0; i < 23; i++)
575 sr = (double) BLOSUM62[i][seqNum] + 4;
577 // Calculate X with another loop over residues
578 // System.out.println("Xi " + i + " " + x[i] + " " + sr);
581 tot += (xx[i] * xx[i]);
584 bigtot += Math.sqrt(tot);
587 // This is the quality for one column
593 // bigtot = bigtot * (size-cons2[j][23])/size;
594 quality.addElement(new Double(bigtot));
596 // Need to normalize by gaps
599 double newmax = -10000;
601 for (j = startRes; j <= endRes; j++)
603 tmp = quality.elementAt(j).doubleValue();
604 tmp = ((max - tmp) * (size - cons2[j][23])) / size;
606 // System.out.println(tmp+ " " + j);
607 quality.setElementAt(new Double(tmp), j);
615 // System.out.println("Quality " + s);
616 qualityRange[0] = 0D;
617 qualityRange[1] = newmax;
621 * Complete the given consensus and quuality annotation rows. Note: currently
622 * this method will enlarge the given annotation row if it is too small,
623 * otherwise will leave its length unchanged.
625 * @param conservation
626 * conservation annotation row
628 * (optional - may be null)
630 * first column for conservation
632 * extent of conservation
634 public void completeAnnotations(AlignmentAnnotation conservation,
635 AlignmentAnnotation quality2, int istart, int alWidth)
637 char[] sequence = getConsSequence().getSequence();
649 maxB = 0f - minB; // scalable range for colouring both Conservation and
659 if (conservation != null && conservation.annotations != null
660 && conservation.annotations.length < alWidth)
662 conservation.annotations = new Annotation[alWidth];
665 if (quality2 != null)
667 quality2.graphMax = (float) qualityRange[1];
668 if (quality2.annotations != null
669 && quality2.annotations.length < alWidth)
671 quality2.annotations = new Annotation[alWidth];
673 qmin = (float) qualityRange[0];
674 qmax = (float) qualityRange[1];
677 for (int i = istart; i < alWidth; i++)
683 if (Character.isDigit(c))
696 if (conservation != null)
698 float vprop = value - min;
700 int consp = i - start;
701 String conssym = (value > 0 && consp > -1 && consp < consSymbs.length) ? consSymbs[consp]
703 conservation.annotations[i] = new Annotation(String.valueOf(c),
704 conssym, ' ', value, new Color(minR + (maxR * vprop), minG
705 + (maxG * vprop), minB + (maxB * vprop)));
709 if (quality2 != null)
711 value = quality.elementAt(i).floatValue();
712 float 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);