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.analysis.ResidueCount.SymbolCounts;
24 import jalview.datamodel.AlignedCodonFrame;
25 import jalview.datamodel.AlignmentAnnotation;
26 import jalview.datamodel.AlignmentI;
27 import jalview.datamodel.Annotation;
28 import jalview.datamodel.SequenceI;
29 import jalview.ext.android.SparseIntArray;
30 import jalview.util.Comparison;
31 import jalview.util.Format;
32 import jalview.util.MappingUtils;
33 import jalview.util.QuickSort;
35 import java.util.Arrays;
36 import java.util.Hashtable;
37 import java.util.List;
40 * Takes in a vector or array of sequences and column start and column end and
41 * returns a new Hashtable[] of size maxSeqLength, if Hashtable not supplied.
42 * This class is used extensively in calculating alignment colourschemes that
43 * depend on the amount of conservation in each alignment column.
48 public class AAFrequency
50 public static final String MAXCOUNT = "C";
52 public static final String MAXRESIDUE = "R";
54 public static final String PID_GAPS = "G";
56 public static final String PID_NOGAPS = "N";
58 public static final String PROFILE = "P";
60 public static final String ENCODED_CHARS = "E";
63 * Quick look-up of String value of char 'A' to 'Z'
65 private static final String[] CHARS = new String['Z' - 'A' + 1];
69 for (char c = 'A'; c <= 'Z'; c++)
71 CHARS[c - 'A'] = String.valueOf(c);
75 public static final Profile[] calculate(List<SequenceI> list,
78 return calculate(list, start, end, false);
81 public static final Profile[] calculate(List<SequenceI> sequences,
82 int start, int end, boolean profile)
84 SequenceI[] seqs = new SequenceI[sequences.size()];
86 synchronized (sequences)
88 for (int i = 0; i < sequences.size(); i++)
90 seqs[i] = sequences.get(i);
91 if (seqs[i].getLength() > width)
93 width = seqs[i].getLength();
97 Profile[] reply = new Profile[width];
104 calculate(seqs, start, end, reply, profile);
110 * Calculate the consensus symbol(s) for each column in the given range.
114 * start column (inclusive, base zero)
116 * end column (exclusive)
118 * array in which to store profile per column
119 * @param saveFullProfile
120 * if true, store all symbol counts
122 public static final void calculate(final SequenceI[] sequences,
123 int start, int end, Profile[] result, boolean saveFullProfile)
125 // long now = System.currentTimeMillis();
126 int seqCount = sequences.length;
127 boolean nucleotide = false;
128 int nucleotideCount = 0;
129 int peptideCount = 0;
131 for (int column = start; column < end; column++)
134 * Apply a heuristic to detect nucleotide data (which can
135 * be counted in more compact arrays); here we test for
136 * more than 90% nucleotide; recheck every 10 columns in case
137 * of misleading data e.g. highly conserved Alanine in peptide!
138 * Mistakenly guessing nucleotide has a small performance cost,
139 * as it will result in counting in sparse arrays.
140 * Mistakenly guessing peptide has a small space cost,
141 * as it will use a larger than necessary array to hold counts.
143 if (nucleotideCount > 100 && column % 10 == 0)
145 nucleotide = (9 * peptideCount < nucleotideCount);
147 ResidueCount residueCounts = new ResidueCount(nucleotide);
149 for (int row = 0; row < seqCount; row++)
151 if (sequences[row] == null)
154 .println("WARNING: Consensus skipping null sequence - possible race condition.");
157 char[] seq = sequences[row].getSequence();
158 if (seq.length > column)
160 char c = seq[column];
161 residueCounts.add(c);
162 if (Comparison.isNucleotide(c))
166 else if (!Comparison.isGap(c))
174 * count a gap if the sequence doesn't reach this column
176 residueCounts.addGap();
180 int maxCount = residueCounts.getModalCount();
181 String maxResidue = residueCounts.getResiduesForCount(maxCount);
182 int gapCount = residueCounts.getGapCount();
183 Profile profile = new Profile(seqCount, gapCount, maxCount,
188 profile.setCounts(residueCounts);
191 result[column] = profile;
193 // long elapsed = System.currentTimeMillis() - now;
194 // System.out.println(elapsed);
198 * Make an estimate of the profile size we are going to compute i.e. how many
199 * different characters may be present in it. Overestimating has a cost of
200 * using more memory than necessary. Underestimating has a cost of needing to
201 * extend the SparseIntArray holding the profile counts.
203 * @param profileSizes
204 * counts of sizes of profiles so far encountered
207 static int estimateProfileSize(SparseIntArray profileSizes)
209 if (profileSizes.size() == 0)
215 * could do a statistical heuristic here e.g. 75%ile
216 * for now just return the largest value
218 return profileSizes.keyAt(profileSizes.size() - 1);
222 * Derive the consensus annotations to be added to the alignment for display.
223 * This does not recompute the raw data, but may be called on a change in
224 * display options, such as 'ignore gaps', which may in turn result in a
225 * change in the derived values.
228 * the annotation row to add annotations to
230 * the source consensus data
236 * if true, normalise residue percentages ignoring gaps
237 * @param showSequenceLogo
238 * if true include all consensus symbols, else just show modal
241 * number of sequences
243 public static void completeConsensus(AlignmentAnnotation consensus,
244 Profile[] profiles, int iStart, int width, boolean ignoreGaps,
245 boolean showSequenceLogo, long nseq)
247 // long now = System.currentTimeMillis();
248 if (consensus == null || consensus.annotations == null
249 || consensus.annotations.length < width)
252 * called with a bad alignment annotation row
253 * wait for it to be initialised properly
258 final int dp = getPercentageDp(nseq);
260 for (int i = iStart; i < width; i++)
263 if (i >= profiles.length || ((profile = profiles[i]) == null))
266 * happens if sequences calculated over were
267 * shorter than alignment width
269 consensus.annotations[i] = null;
273 float value = profile.getPercentageIdentity(ignoreGaps);
275 String description = getTooltip(profile, value, showSequenceLogo,
278 consensus.annotations[i] = new Annotation(profile.getModalResidue(),
279 description, ' ', value);
281 // long elapsed = System.currentTimeMillis() - now;
282 // System.out.println(-elapsed);
286 * Returns a tooltip showing either
288 * <li>the full profile (percentages of all residues present), if
289 * showSequenceLogo is true, or</li>
290 * <li>just the modal (most common) residue(s), if showSequenceLogo is false</li>
292 * Percentages are as a fraction of all sequence, or only ungapped sequences
293 * if ignoreGaps is true.
297 * @param showSequenceLogo
300 * the number of decimal places to format percentages to
303 static String getTooltip(Profile profile, float pid,
304 boolean showSequenceLogo, boolean ignoreGaps, int dp)
306 ResidueCount counts = profile.getCounts();
308 String description = null;
309 if (counts != null && showSequenceLogo)
311 int normaliseBy = ignoreGaps ? profile.getNonGapped() : profile
313 description = counts.getTooltip(normaliseBy, dp);
317 StringBuilder sb = new StringBuilder(64);
318 String maxRes = profile.getModalResidue();
319 if (maxRes.length() > 1)
321 sb.append("[").append(maxRes).append("] ");
326 sb.append(maxRes).append(" ");
328 Format.appendPercentage(sb, pid, dp);
330 description = sb.toString();
336 * Returns the sorted profile for the given consensus data. The returned array
340 * [profileType, numberOfValues, nonGapCount, charValue1, percentage1, charValue2, percentage2, ...]
341 * in descending order of percentage value
345 * the data object from which to extract and sort values
347 * if true, only non-gapped values are included in percentage
351 public static int[] extractProfile(Profile profile,
354 int[] rtnval = new int[64];
355 ResidueCount counts = profile.getCounts();
361 SymbolCounts symbolCounts = counts.getSymbolCounts();
362 char[] symbols = symbolCounts.symbols;
363 int[] values = symbolCounts.values;
364 QuickSort.sort(values, symbols);
365 int nextArrayPos = 2;
366 int totalPercentage = 0;
367 final int divisor = ignoreGaps ? profile.getNonGapped() : profile
371 * traverse the arrays in reverse order (highest counts first)
373 for (int i = symbols.length - 1; i >= 0; i--)
375 int theChar = symbols[i];
376 int charCount = values[i];
378 rtnval[nextArrayPos++] = theChar;
379 final int percentage = (charCount * 100) / divisor;
380 rtnval[nextArrayPos++] = percentage;
381 totalPercentage += percentage;
383 rtnval[0] = symbols.length;
384 rtnval[1] = totalPercentage;
385 int[] result = new int[rtnval.length + 1];
386 result[0] = AlignmentAnnotation.SEQUENCE_PROFILE;
387 System.arraycopy(rtnval, 0, result, 1, rtnval.length);
393 * Extract a sorted extract of cDNA codon profile data. The returned array
397 * [profileType, numberOfValues, totalCount, charValue1, percentage1, charValue2, percentage2, ...]
398 * in descending order of percentage value, where the character values encode codon triplets
404 public static int[] extractCdnaProfile(Hashtable hashtable,
407 // this holds #seqs, #ungapped, and then codon count, indexed by encoded
409 int[] codonCounts = (int[]) hashtable.get(PROFILE);
410 int[] sortedCounts = new int[codonCounts.length - 2];
411 System.arraycopy(codonCounts, 2, sortedCounts, 0,
412 codonCounts.length - 2);
414 int[] result = new int[3 + 2 * sortedCounts.length];
415 // first value is just the type of profile data
416 result[0] = AlignmentAnnotation.CDNA_PROFILE;
418 char[] codons = new char[sortedCounts.length];
419 for (int i = 0; i < codons.length; i++)
421 codons[i] = (char) i;
423 QuickSort.sort(sortedCounts, codons);
424 int totalPercentage = 0;
425 int distinctValuesCount = 0;
427 int divisor = ignoreGaps ? codonCounts[1] : codonCounts[0];
428 for (int i = codons.length - 1; i >= 0; i--)
430 final int codonCount = sortedCounts[i];
433 break; // nothing else of interest here
435 distinctValuesCount++;
436 result[j++] = codons[i];
437 final int percentage = codonCount * 100 / divisor;
438 result[j++] = percentage;
439 totalPercentage += percentage;
441 result[2] = totalPercentage;
444 * Just return the non-zero values
446 // todo next value is redundant if we limit the array to non-zero counts
447 result[1] = distinctValuesCount;
448 return Arrays.copyOfRange(result, 0, j);
452 * Compute a consensus for the cDNA coding for a protein alignment.
455 * the protein alignment (which should hold mappings to cDNA
458 * the consensus data stores to be populated (one per column)
460 public static void calculateCdna(AlignmentI alignment,
461 Hashtable[] hconsensus)
463 final char gapCharacter = alignment.getGapCharacter();
464 List<AlignedCodonFrame> mappings = alignment.getCodonFrames();
465 if (mappings == null || mappings.isEmpty())
470 int cols = alignment.getWidth();
471 for (int col = 0; col < cols; col++)
473 // todo would prefer a Java bean for consensus data
474 Hashtable<String, int[]> columnHash = new Hashtable<String, int[]>();
475 // #seqs, #ungapped seqs, counts indexed by (codon encoded + 1)
476 int[] codonCounts = new int[66];
477 codonCounts[0] = alignment.getSequences().size();
478 int ungappedCount = 0;
479 for (SequenceI seq : alignment.getSequences())
481 if (seq.getCharAt(col) == gapCharacter)
485 List<char[]> codons = MappingUtils
486 .findCodonsFor(seq, col, mappings);
487 for (char[] codon : codons)
489 int codonEncoded = CodingUtils.encodeCodon(codon);
490 if (codonEncoded >= 0)
492 codonCounts[codonEncoded + 2]++;
497 codonCounts[1] = ungappedCount;
498 // todo: sort values here, save counts and codons?
499 columnHash.put(PROFILE, codonCounts);
500 hconsensus[col] = columnHash;
505 * Derive displayable cDNA consensus annotation from computed consensus data.
507 * @param consensusAnnotation
508 * the annotation row to be populated for display
509 * @param consensusData
510 * the computed consensus data
511 * @param showProfileLogo
512 * if true show all symbols present at each position, else only the
515 * the number of sequences in the alignment
517 public static void completeCdnaConsensus(
518 AlignmentAnnotation consensusAnnotation,
519 Hashtable[] consensusData, boolean showProfileLogo, int nseqs)
521 if (consensusAnnotation == null
522 || consensusAnnotation.annotations == null
523 || consensusAnnotation.annotations.length < consensusData.length)
525 // called with a bad alignment annotation row - wait for it to be
526 // initialised properly
530 // ensure codon triplet scales with font size
531 consensusAnnotation.scaleColLabel = true;
532 for (int col = 0; col < consensusData.length; col++)
534 Hashtable hci = consensusData[col];
537 // gapped protein column?
540 // array holds #seqs, #ungapped, then codon counts indexed by codon
541 final int[] codonCounts = (int[]) hci.get(PROFILE);
545 * First pass - get total count and find the highest
547 final char[] codons = new char[codonCounts.length - 2];
548 for (int j = 2; j < codonCounts.length; j++)
550 final int codonCount = codonCounts[j];
551 codons[j - 2] = (char) (j - 2);
552 totalCount += codonCount;
556 * Sort array of encoded codons by count ascending - so the modal value
557 * goes to the end; start by copying the count (dropping the first value)
559 int[] sortedCodonCounts = new int[codonCounts.length - 2];
560 System.arraycopy(codonCounts, 2, sortedCodonCounts, 0,
561 codonCounts.length - 2);
562 QuickSort.sort(sortedCodonCounts, codons);
564 int modalCodonEncoded = codons[codons.length - 1];
565 int modalCodonCount = sortedCodonCounts[codons.length - 1];
566 String modalCodon = String.valueOf(CodingUtils
567 .decodeCodon(modalCodonEncoded));
568 if (sortedCodonCounts.length > 1
569 && sortedCodonCounts[codons.length - 2] == sortedCodonCounts[codons.length - 1])
572 * two or more codons share the modal count
576 float pid = sortedCodonCounts[sortedCodonCounts.length - 1] * 100
577 / (float) totalCount;
580 * todo ? Replace consensus hashtable with sorted arrays of codons and
581 * counts (non-zero only). Include total count in count array [0].
585 * Scan sorted array backwards for most frequent values first. Show
586 * repeated values compactly.
588 StringBuilder mouseOver = new StringBuilder(32);
589 StringBuilder samePercent = new StringBuilder();
590 String percent = null;
591 String lastPercent = null;
592 int percentDecPl = getPercentageDp(nseqs);
594 for (int j = codons.length - 1; j >= 0; j--)
596 int codonCount = sortedCodonCounts[j];
600 * remaining codons are 0% - ignore, but finish off the last one if
603 if (samePercent.length() > 0)
605 mouseOver.append(samePercent).append(": ").append(percent)
610 int codonEncoded = codons[j];
611 final int pct = codonCount * 100 / totalCount;
612 String codon = String
613 .valueOf(CodingUtils.decodeCodon(codonEncoded));
614 StringBuilder sb = new StringBuilder();
615 Format.appendPercentage(sb, pct, percentDecPl);
616 percent = sb.toString();
617 if (showProfileLogo || codonCount == modalCodonCount)
619 if (percent.equals(lastPercent) && j > 0)
621 samePercent.append(samePercent.length() == 0 ? "" : ", ");
622 samePercent.append(codon);
626 if (samePercent.length() > 0)
628 mouseOver.append(samePercent).append(": ")
629 .append(lastPercent).append("% ");
631 samePercent.setLength(0);
632 samePercent.append(codon);
634 lastPercent = percent;
638 consensusAnnotation.annotations[col] = new Annotation(modalCodon,
639 mouseOver.toString(), ' ', pid);
644 * Returns the number of decimal places to show for profile percentages. For
645 * less than 100 sequences, returns zero (the integer percentage value will be
646 * displayed). For 100-999 sequences, returns 1, for 1000-9999 returns 2, etc.
651 protected static int getPercentageDp(long nseq)