X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FStructureFrequency.java;h=cc97be092aeb2ffcc1ab9f98554a96224e879368;hb=a83adb45bdf9554e270921b4baad94defd314b36;hp=29d02fe6168b562302db186b5f0c3af625db99fa;hpb=0715e493687eeb93d3b3f39cd1ff25e2671ce0e9;p=jalview.git diff --git a/src/jalview/analysis/StructureFrequency.java b/src/jalview/analysis/StructureFrequency.java index 29d02fe..cc97be0 100644 --- a/src/jalview/analysis/StructureFrequency.java +++ b/src/jalview/analysis/StructureFrequency.java @@ -24,9 +24,9 @@ import jalview.datamodel.AlignmentAnnotation; import jalview.datamodel.Annotation; import jalview.datamodel.SequenceFeature; import jalview.datamodel.SequenceI; +import jalview.util.Comparison; import jalview.util.Format; -import java.util.ArrayList; import java.util.Hashtable; /** @@ -93,11 +93,11 @@ public class StructureFrequency * @param rnaStruc */ public static final void calculate(SequenceI[] sequences, int start, - int end, Hashtable[] result, boolean profile, + int end, Hashtable[] result, boolean profile, AlignmentAnnotation rnaStruc) { - Hashtable residueHash; + Hashtable residueHash; String maxResidue; char[] struc = rnaStruc.getRNAStruc().toCharArray(); @@ -111,8 +111,9 @@ public class StructureFrequency for (int i = start; i < end; i++) // foreach column { - int canonicalOrWobblePairCount = 0; + int canonicalOrWobblePairCount = 0, canonical = 0; int otherPairCount = 0; + int nongap = 0; maxResidue = "-"; values = new int[255]; pairs = new int[255][255]; @@ -147,25 +148,20 @@ public class StructureFrequency { if (sequences[j] == null) { - System.err - .println("WARNING: Consensus skipping null sequence - possible race condition."); + System.err.println( + "WARNING: Consensus skipping null sequence - possible race condition."); continue; } - c = sequences[j].getCharAt(i); - // standard representation for gaps in sequence and structure - if (c == '.' || c == ' ') - { - c = '-'; - } + c = sequences[j].getCharAt(i); + cEnd = sequences[j].getCharAt(bpEnd); - if (c == '-') + if (Comparison.isGap(c) || Comparison.isGap(cEnd)) { values['-']++; continue; } - cEnd = sequences[j].getCharAt(bpEnd); - + nongap++; /* * ensure upper-case for counting purposes */ @@ -179,36 +175,56 @@ public class StructureFrequency } if (Rna.isCanonicalOrWobblePair(c, cEnd)) { - values['(']++; - maxResidue = "("; canonicalOrWobblePairCount++; + if (Rna.isCanonicalPair(c, cEnd)) + { + canonical++; + } } else { - values['[']++; - maxResidue = "["; otherPairCount++; } pairs[c][cEnd]++; } } - // nonGap++; } - residueHash = new Hashtable(); + residueHash = new Hashtable<>(); if (profile) { // TODO 1-dim array with jsize in [0], nongapped in [1]; or Pojo - residueHash.put(PROFILE, new int[][] { values, - new int[] { jSize, (jSize - values['-']) } }); + residueHash.put(PROFILE, + new int[][] + { values, new int[] { jSize, (jSize - values['-']) } }); residueHash.put(PAIRPROFILE, pairs); } - int count = Math.max(canonicalOrWobblePairCount, otherPairCount); - if (!maxResidue.equals("-")) + values['('] = canonicalOrWobblePairCount; + values['['] = canonical; + values['{'] = otherPairCount; + /* + * the count is the number of valid pairs (as a percentage, determines + * the relative size of the profile logo) + */ + int count = canonicalOrWobblePairCount; + + /* + * display '(' if most pairs are canonical, or as + * '[' if there are more wobble pairs. + */ + if (canonicalOrWobblePairCount > 0 || otherPairCount > 0) { - maxResidue = canonicalOrWobblePairCount >= otherPairCount ? "(" - : "["; + if (canonicalOrWobblePairCount >= otherPairCount) + { + maxResidue = (canonicalOrWobblePairCount - canonical) < canonical + ? "(" + : "["; + } + else + { + maxResidue = "{"; + } } residueHash.put(MAXCOUNT, new Integer(count)); residueHash.put(MAXRESIDUE, maxResidue); @@ -216,8 +232,9 @@ public class StructureFrequency percentage = ((float) count * 100) / jSize; residueHash.put(PID_GAPS, new Float(percentage)); - // percentage = ((float) count * 100) / (float) nongap; - // residueHash.put(PID_NOGAPS, new Float(percentage)); + percentage = ((float) count * 100) / nongap; + residueHash.put(PID_NOGAPS, new Float(percentage)); + if (result[i] == null) { result[i] = residueHash; @@ -226,15 +243,19 @@ public class StructureFrequency { values[')'] = values['(']; values[']'] = values['[']; + values['}'] = values['{']; values['('] = 0; values['['] = 0; - maxResidue = maxResidue.equals("(") ? ")" : "]"; + values['{'] = 0; + maxResidue = maxResidue.equals("(") ? ")" + : maxResidue.equals("[") ? "]" : "}"; - residueHash = new Hashtable(); + residueHash = new Hashtable<>(); if (profile) { - residueHash.put(PROFILE, new int[][] { values, - new int[] { jSize, (jSize - values['-']) } }); + residueHash.put(PROFILE, + new int[][] + { values, new int[] { jSize, (jSize - values['-']) } }); residueHash.put(PAIRPROFILE, pairs); } @@ -245,6 +266,9 @@ public class StructureFrequency percentage = ((float) count * 100) / jSize; residueHash.put(PID_GAPS, new Float(percentage)); + percentage = ((float) count * 100) / nongap; + residueHash.put(PID_NOGAPS, new Float(percentage)); + result[bpEnd] = residueHash; } } @@ -263,7 +287,7 @@ public class StructureFrequency * @param includeAllConsSymbols */ public static void completeConsensus(AlignmentAnnotation consensus, - Hashtable[] hconsensus, int iStart, int width, + Hashtable[] hconsensus, int iStart, int width, boolean ignoreGapsInConsensusCalculation, boolean includeAllConsSymbols, long nseq) { @@ -290,7 +314,7 @@ public class StructureFrequency for (int i = iStart; i < width; i++) { - Hashtable hci; + Hashtable hci; if (i >= hconsensus.length || ((hci = hconsensus[i]) == null)) { // happens if sequences calculated over were shorter than alignment @@ -390,7 +414,7 @@ public class StructureFrequency * @param hconsensus * @return profile of the given column */ - public static int[] extractProfile(Hashtable hconsensus, + public static int[] extractProfile(Hashtable hconsensus, boolean ignoreGapsInConsensusCalculation) { int[] rtnval = new int[STRUCTURE_PROFILE_LENGTH]; // 2*(5*5)+2 @@ -422,7 +446,8 @@ public class StructureFrequency int valuesCount = 0; rtnval[1] = 0; int offset = 2; - final int divisor = profile[1][ignoreGapsInConsensusCalculation ? 1 : 0]; + final int divisor = profile[1][ignoreGapsInConsensusCalculation ? 1 + : 0]; for (int c = 624; c > 0; c--) { if (vl[c] > 0) @@ -442,23 +467,4 @@ public class StructureFrequency System.arraycopy(rtnval, 0, result, 1, rtnval.length); return result; } - - public static void main(String args[]) - { - // Short test to see if checkBpType works - ArrayList test = new ArrayList(); - test.add("A"); - test.add("c"); - test.add("g"); - test.add("T"); - test.add("U"); - for (String i : test) - { - for (String j : test) - { - System.out.println(i + "-" + j + ": " - + Rna.isCanonicalOrWobblePair(i.charAt(0), j.charAt(0))); - } - } - } }