X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FAAFrequency.java;h=52dedcfba9ae9acc265c3d928675a1c09381c574;hb=57381bb294ebc639073c198d14b03e16ca6aff97;hp=61e11c4ec70f9da32b93ae5259cae2f4f7ce4019;hpb=d94db336381e48d9155736ca4ea058d9f747439c;p=jalview.git diff --git a/src/jalview/analysis/AAFrequency.java b/src/jalview/analysis/AAFrequency.java index 61e11c4..52dedcf 100755 --- a/src/jalview/analysis/AAFrequency.java +++ b/src/jalview/analysis/AAFrequency.java @@ -20,11 +20,16 @@ */ package jalview.analysis; -import jalview.analysis.ResidueCount.SymbolCounts; import jalview.datamodel.AlignedCodonFrame; import jalview.datamodel.AlignmentAnnotation; import jalview.datamodel.AlignmentI; import jalview.datamodel.Annotation; +import jalview.datamodel.Profile; +import jalview.datamodel.ProfileI; +import jalview.datamodel.Profiles; +import jalview.datamodel.ProfilesI; +import jalview.datamodel.ResidueCount; +import jalview.datamodel.ResidueCount.SymbolCounts; import jalview.datamodel.SequenceI; import jalview.ext.android.SparseIntArray; import jalview.util.Comparison; @@ -32,6 +37,7 @@ import jalview.util.Format; import jalview.util.MappingUtils; import jalview.util.QuickSort; +import java.awt.Color; import java.util.Arrays; import java.util.Hashtable; import java.util.List; @@ -62,13 +68,13 @@ public class AAFrequency } } - public static final Profile[] calculate(List list, - int start, int end) + public static final ProfilesI calculate(List list, int start, + int end) { return calculate(list, start, end, false); } - public static final Profile[] calculate(List sequences, + public static final ProfilesI calculate(List sequences, int start, int end, boolean profile) { SequenceI[] seqs = new SequenceI[sequences.size()]; @@ -78,20 +84,19 @@ public class AAFrequency for (int i = 0; i < sequences.size(); i++) { seqs[i] = sequences.get(i); - if (seqs[i].getLength() > width) + int length = seqs[i].getLength(); + if (length > width) { - width = seqs[i].getLength(); + width = length; } } - Profile[] reply = new Profile[width]; - if (end >= width) { end = width; } - calculate(seqs, start, end, reply, profile); + ProfilesI reply = calculate(seqs, width, start, end, profile); return reply; } } @@ -100,17 +105,17 @@ public class AAFrequency * Calculate the consensus symbol(s) for each column in the given range. * * @param sequences + * @param width + * the full width of the alignment * @param start * start column (inclusive, base zero) * @param end * end column (exclusive) - * @param result - * array in which to store profile per column * @param saveFullProfile * if true, store all symbol counts */ - public static final void calculate(final SequenceI[] sequences, - int start, int end, Profile[] result, boolean saveFullProfile) + public static final ProfilesI calculate(final SequenceI[] sequences, + int width, int start, int end, boolean saveFullProfile) { // long now = System.currentTimeMillis(); int seqCount = sequences.length; @@ -118,6 +123,8 @@ public class AAFrequency int nucleotideCount = 0; int peptideCount = 0; + ProfileI[] result = new ProfileI[width]; + for (int column = start; column < end; column++) { /* @@ -140,14 +147,13 @@ public class AAFrequency { if (sequences[row] == null) { - System.err - .println("WARNING: Consensus skipping null sequence - possible race condition."); + System.err.println( + "WARNING: Consensus skipping null sequence - possible race condition."); continue; } - char[] seq = sequences[row].getSequence(); - if (seq.length > column) + if (sequences[row].getLength() > column) { - char c = seq[column]; + char c = sequences[row].getCharAt(column); residueCounts.add(c); if (Comparison.isNucleotide(c)) { @@ -170,7 +176,7 @@ public class AAFrequency int maxCount = residueCounts.getModalCount(); String maxResidue = residueCounts.getResiduesForCount(maxCount); int gapCount = residueCounts.getGapCount(); - Profile profile = new Profile(seqCount, gapCount, maxCount, + ProfileI profile = new Profile(seqCount, gapCount, maxCount, maxResidue); if (saveFullProfile) @@ -180,6 +186,7 @@ public class AAFrequency result[column] = profile; } + return new Profiles(result); // long elapsed = System.currentTimeMillis() - now; // System.out.println(elapsed); } @@ -218,10 +225,10 @@ public class AAFrequency * the annotation row to add annotations to * @param profiles * the source consensus data - * @param iStart - * start column - * @param width - * end column + * @param startCol + * start column (inclusive) + * @param endCol + * end column (exclusive) * @param ignoreGaps * if true, normalise residue percentages ignoring gaps * @param showSequenceLogo @@ -231,12 +238,12 @@ public class AAFrequency * number of sequences */ public static void completeConsensus(AlignmentAnnotation consensus, - Profile[] profiles, int iStart, int width, boolean ignoreGaps, + ProfilesI profiles, int startCol, int endCol, boolean ignoreGaps, boolean showSequenceLogo, long nseq) { // long now = System.currentTimeMillis(); if (consensus == null || consensus.annotations == null - || consensus.annotations.length < width) + || consensus.annotations.length < endCol) { /* * called with a bad alignment annotation row @@ -245,21 +252,21 @@ public class AAFrequency return; } - final int dp = getPercentageDp(nseq); - - for (int i = iStart; i < width; i++) + for (int i = startCol; i < endCol; i++) { - Profile profile; - if (i >= profiles.length || ((profile = profiles[i]) == null)) + ProfileI profile = profiles.get(i); + if (profile == null) { /* * happens if sequences calculated over were * shorter than alignment width */ consensus.annotations[i] = null; - continue; + return; } + final int dp = getPercentageDp(nseq); + float value = profile.getPercentageIdentity(ignoreGaps); String description = getTooltip(profile, value, showSequenceLogo, @@ -274,19 +281,71 @@ public class AAFrequency { modalResidue = "+"; } - consensus.annotations[i] = new Annotation(modalResidue, - description, ' ', value); + consensus.annotations[i] = new Annotation(modalResidue, description, + ' ', value); } // long elapsed = System.currentTimeMillis() - now; // System.out.println(-elapsed); } /** + * Derive the gap count annotation row. + * + * @param gaprow + * the annotation row to add annotations to + * @param profiles + * the source consensus data + * @param startCol + * start column (inclusive) + * @param endCol + * end column (exclusive) + */ + public static void completeGapAnnot(AlignmentAnnotation gaprow, + ProfilesI profiles, int startCol, int endCol, long nseq) + { + if (gaprow == null || gaprow.annotations == null + || gaprow.annotations.length < endCol) + { + /* + * called with a bad alignment annotation row + * wait for it to be initialised properly + */ + return; + } + // always set ranges again + gaprow.graphMax = nseq; + gaprow.graphMin = 0; + double scale = 0.8 / nseq; + for (int i = startCol; i < endCol; i++) + { + ProfileI profile = profiles.get(i); + if (profile == null) + { + /* + * happens if sequences calculated over were + * shorter than alignment width + */ + gaprow.annotations[i] = null; + return; + } + + final int gapped = profile.getNonGapped(); + + String description = "" + gapped; + + gaprow.annotations[i] = new Annotation("", description, '\0', gapped, + jalview.util.ColorUtils.bleachColour(Color.DARK_GRAY, + (float) scale * gapped)); + } + } + + /** * Returns a tooltip showing either *
    *
  • the full profile (percentages of all residues present), if * showSequenceLogo is true, or
    • - *
  • just the modal (most common) residue(s), if showSequenceLogo is false
    • + *
  • just the modal (most common) residue(s), if showSequenceLogo is + * false
    • *
* Percentages are as a fraction of all sequence, or only ungapped sequences * if ignoreGaps is true. @@ -299,7 +358,7 @@ public class AAFrequency * the number of decimal places to format percentages to * @return */ - static String getTooltip(Profile profile, float pid, + static String getTooltip(ProfileI profile, float pid, boolean showSequenceLogo, boolean ignoreGaps, int dp) { ResidueCount counts = profile.getCounts(); @@ -307,8 +366,8 @@ public class AAFrequency String description = null; if (counts != null && showSequenceLogo) { - int normaliseBy = ignoreGaps ? profile.getNonGapped() : profile - .getHeight(); + int normaliseBy = ignoreGaps ? profile.getNonGapped() + : profile.getHeight(); description = counts.getTooltip(normaliseBy, dp); } else @@ -339,7 +398,7 @@ public class AAFrequency * contains * * 
-   *    [profileType, numberOfValues, nonGapCount, charValue1, percentage1, charValue2, percentage2, ...]
+   *    [profileType, numberOfValues, totalPercent, charValue1, percentage1, charValue2, percentage2, ...]
    * in descending order of percentage value
    *
* @@ -350,10 +409,8 @@ public class AAFrequency * calculations * @return */ - public static int[] extractProfile(Profile profile, - boolean ignoreGaps) + public static int[] extractProfile(ProfileI profile, boolean ignoreGaps) { - int[] rtnval = new int[64]; ResidueCount counts = profile.getCounts(); if (counts == null) { @@ -364,29 +421,51 @@ public class AAFrequency char[] symbols = symbolCounts.symbols; int[] values = symbolCounts.values; QuickSort.sort(values, symbols); - int nextArrayPos = 2; int totalPercentage = 0; - final int divisor = ignoreGaps ? profile.getNonGapped() : profile - .getHeight(); + final int divisor = ignoreGaps ? profile.getNonGapped() + : profile.getHeight(); /* * traverse the arrays in reverse order (highest counts first) */ + int[] result = new int[3 + 2 * symbols.length]; + int nextArrayPos = 3; + int nonZeroCount = 0; + for (int i = symbols.length - 1; i >= 0; i--) { int theChar = symbols[i]; int charCount = values[i]; - - rtnval[nextArrayPos++] = theChar; final int percentage = (charCount * 100) / divisor; - rtnval[nextArrayPos++] = percentage; + if (percentage == 0) + { + /* + * this count (and any remaining) round down to 0% - discard + */ + break; + } + nonZeroCount++; + result[nextArrayPos++] = theChar; + result[nextArrayPos++] = percentage; totalPercentage += percentage; } - rtnval[0] = symbols.length; - rtnval[1] = totalPercentage; - int[] result = new int[rtnval.length + 1]; + + /* + * truncate array if any zero values were discarded + */ + if (nonZeroCount < symbols.length) + { + int[] tmp = new int[3 + 2 * nonZeroCount]; + System.arraycopy(result, 0, tmp, 0, tmp.length); + result = tmp; + } + + /* + * fill in 'header' values + */ result[0] = AlignmentAnnotation.SEQUENCE_PROFILE; - System.arraycopy(rtnval, 0, result, 1, rtnval.length); + result[1] = nonZeroCount; + result[2] = totalPercentage; return result; } @@ -396,14 +475,15 @@ public class AAFrequency * contains * * 
-   *    [profileType, numberOfValues, totalCount, charValue1, percentage1, charValue2, percentage2, ...]
+   *    [profileType, numberOfValues, totalPercentage, charValue1, percentage1, charValue2, percentage2, ...]
    * in descending order of percentage value, where the character values encode codon triplets
    *
* * @param hashtable * @return */ - public static int[] extractCdnaProfile(Hashtable hashtable, + public static int[] extractCdnaProfile( + Hashtable hashtable, boolean ignoreGaps) { // this holds #seqs, #ungapped, and then codon count, indexed by encoded @@ -434,9 +514,16 @@ public class AAFrequency { break; // nothing else of interest here } + final int percentage = codonCount * 100 / divisor; + if (percentage == 0) + { + /* + * this (and any remaining) values rounded down to 0 - discard + */ + break; + } distinctValuesCount++; result[j++] = codons[i]; - final int percentage = codonCount * 100 / divisor; result[j++] = percentage; totalPercentage += percentage; } @@ -460,7 +547,7 @@ public class AAFrequency * the consensus data stores to be populated (one per column) */ public static void calculateCdna(AlignmentI alignment, - Hashtable[] hconsensus) + Hashtable[] hconsensus) { final char gapCharacter = alignment.getGapCharacter(); List mappings = alignment.getCodonFrames(); @@ -473,7 +560,7 @@ public class AAFrequency for (int col = 0; col < cols; col++) { // todo would prefer a Java bean for consensus data - Hashtable columnHash = new Hashtable(); + Hashtable columnHash = new Hashtable<>(); // #seqs, #ungapped seqs, counts indexed by (codon encoded + 1) int[] codonCounts = new int[66]; codonCounts[0] = alignment.getSequences().size(); @@ -484,8 +571,8 @@ public class AAFrequency { continue; } - List codons = MappingUtils - .findCodonsFor(seq, col, mappings); + List codons = MappingUtils.findCodonsFor(seq, col, + mappings); for (char[] codon : codons) { int codonEncoded = CodingUtils.encodeCodon(codon); @@ -493,6 +580,7 @@ public class AAFrequency { codonCounts[codonEncoded + 2]++; ungappedCount++; + break; } } } @@ -518,7 +606,8 @@ public class AAFrequency */ public static void completeCdnaConsensus( AlignmentAnnotation consensusAnnotation, - Hashtable[] consensusData, boolean showProfileLogo, int nseqs) + Hashtable[] consensusData, boolean showProfileLogo, + int nseqs) { if (consensusAnnotation == null || consensusAnnotation.annotations == null @@ -533,7 +622,7 @@ public class AAFrequency consensusAnnotation.scaleColLabel = true; for (int col = 0; col < consensusData.length; col++) { - Hashtable hci = consensusData[col]; + Hashtable hci = consensusData[col]; if (hci == null) { // gapped protein column? @@ -565,10 +654,10 @@ public class AAFrequency int modalCodonEncoded = codons[codons.length - 1]; int modalCodonCount = sortedCodonCounts[codons.length - 1]; - String modalCodon = String.valueOf(CodingUtils - .decodeCodon(modalCodonEncoded)); - if (sortedCodonCounts.length > 1 - && sortedCodonCounts[codons.length - 2] == sortedCodonCounts[codons.length - 1]) + String modalCodon = String + .valueOf(CodingUtils.decodeCodon(modalCodonEncoded)); + if (sortedCodonCounts.length > 1 && sortedCodonCounts[codons.length + - 2] == sortedCodonCounts[codons.length - 1]) { /* * two or more codons share the modal count @@ -627,8 +716,8 @@ public class AAFrequency { if (samePercent.length() > 0) { - mouseOver.append(samePercent).append(": ") - .append(lastPercent).append("% "); + mouseOver.append(samePercent).append(": ").append(lastPercent) + .append("% "); } samePercent.setLength(0); samePercent.append(codon);