X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FStructureFrequency.java;h=aa1e2774beed6ad0f0ac5d3ef36c2758fd8e76a5;hb=bbfdb203ce3a889600caa52478478b75ab03809f;hp=4d83019dae2bbc159dab708a1fc22c3720cf86b0;hpb=46c34fb5b641a5ec57b5eec0e7bf964a9e72219f;p=jalview.git diff --git a/src/jalview/analysis/StructureFrequency.java b/src/jalview/analysis/StructureFrequency.java index 4d83019..aa1e277 100644 --- a/src/jalview/analysis/StructureFrequency.java +++ b/src/jalview/analysis/StructureFrequency.java @@ -1,6 +1,6 @@ /* - * Jalview - A Sequence Alignment Editor and Viewer (Version 2.6) - * Copyright (C) 2010 J Procter, AM Waterhouse, G Barton, M Clamp, S Searle + * Jalview - A Sequence Alignment Editor and Viewer (Version 2.7) + * Copyright (C) 2011 J Procter, AM Waterhouse, J Engelhardt, LM Lui, G Barton, M Clamp, S Searle * * This file is part of Jalview. * @@ -15,6 +15,7 @@ * * You should have received a copy of the GNU General Public License along with Jalview. If not, see . */ + package jalview.analysis; import java.util.*; @@ -46,143 +47,15 @@ public class StructureFrequency public static final String PAIRPROFILE = "B"; - public static final Hashtable[] calculate(Vector sequences, int start, - int end) - { - return calculate(sequences, start, end, false); - } - - public static final Hashtable[] calculate(Vector sequences, int start, - int end, boolean profile) - { - SequenceI[] seqs = new SequenceI[sequences.size()]; - int width = 0; - for (int i = 0; i < sequences.size(); i++) - { - seqs[i] = (SequenceI) sequences.elementAt(i); - if (seqs[i].getLength() > width) - { - width = seqs[i].getLength(); - } - } - - Hashtable[] reply = new Hashtable[width]; - - if (end >= width) - { - end = width; - } - - calculate(seqs, start, end, reply, profile); - - return reply; - } - - public static final void calculate(SequenceI[] sequences, int start, - int end, Hashtable[] result) - { - calculate(sequences, start, end, result, false); - } - - public static final void calculate(SequenceI[] sequences, int start, - int end, Hashtable[] result, boolean profile) - { - Hashtable residueHash; - int maxCount, nongap, i, j, v, jSize = sequences.length; - String maxResidue; - char c; - float percentage; - - int[] values = new int[255]; - - char[] seq; - - for (i = start; i < end; i++) - { - residueHash = new Hashtable(); - maxCount = 0; - maxResidue = ""; - nongap = 0; - values = new int[255]; - - for (j = 0; j < jSize; j++) - { - if (sequences[j] == null) - { - System.err - .println("WARNING: Consensus skipping null sequence - possible race condition."); - continue; - } - seq = sequences[j].getSequence(); - if (seq.length > i) - { - c = seq[i]; - - if (c == '.' || c == ' ') - { - c = '-'; - } - - if (c == '-') - { - values['-']++; - continue; - } - else if ('a' <= c && c <= 'z') - { - c -= 32; // ('a' - 'A'); - } - - nongap++; - values[c]++; - - } - else - { - values['-']++; - } - } - - for (v = 'A'; v < 'Z'; v++) - { - if (values[v] < 2 || values[v] < maxCount) - { - continue; - } - - if (values[v] > maxCount) - { - maxResidue = String.valueOf((char) v); - } - else if (values[v] == maxCount) - { - maxResidue += String.valueOf((char) v); - } - maxCount = values[v]; - } - - if (maxResidue.length() == 0) - { - maxResidue = "-"; - } - if (profile) - { - residueHash.put(PROFILE, new int[][] - { values, new int[] - { jSize, nongap } }); - } - residueHash.put(MAXCOUNT, new Integer(maxCount)); - residueHash.put(MAXRESIDUE, maxResidue); - - percentage = ((float) maxCount * 100) / (float) jSize; - residueHash.put(PID_GAPS, new Float(percentage)); - - percentage = ((float) maxCount * 100) / (float) nongap; - residueHash.put(PID_NOGAPS, new Float(percentage)); - result[i] = residueHash; - } - } - + /** + * Returns the 3' position of a base pair + * + * @param pairs + * Secondary structure annotation + * @param indice + * 5' position of a base pair + * @return 3' position of a base pair + */ public static int findPair(SequenceFeature[] pairs, int indice) { for (int i = 0; i < pairs.length; i++) @@ -210,7 +83,6 @@ public class StructureFrequency int end, Hashtable[] result, boolean profile, AlignmentAnnotation rnaStruc) { - // System.out.println("StructureFrequency.calculateNEW4"); Hashtable residueHash; String maxResidue; char[] seq, struc = rnaStruc.getRNAStruc().toCharArray(); @@ -228,8 +100,14 @@ public class StructureFrequency values = new int[255]; pairs = new int[255][255]; bpEnd = -1; - - s = struc[i]; + if (i < struc.length) + { + s = struc[i]; + } + else + { + s = '-'; + } if (s == '.' || s == ' ') { s = '-'; @@ -273,13 +151,11 @@ public class StructureFrequency cEnd = seq[bpEnd]; if (checkBpType(c, cEnd)) { - values['H']++; // H means it's a helix (structured) + values['(']++; // H means it's a helix (structured) } - // System.out.println("pair: "+c+","+cEnd); pairs[c][cEnd]++; - // System.out.println("pairs: "+c+","+cEnd+" - "+pairs[c][cEnd]); - maxResidue = "H"; + maxResidue = "("; } } // nonGap++; @@ -287,7 +163,6 @@ public class StructureFrequency // UPDATE this for new values if (profile) { - // System.out.println("profile"); residueHash.put(PROFILE, new int[][] { values, new int[] { jSize, (jSize - values['-']) } }); @@ -295,7 +170,7 @@ public class StructureFrequency residueHash.put(PAIRPROFILE, pairs); } - count = values['H']; + count = values['(']; residueHash.put(MAXCOUNT, new Integer(count)); residueHash.put(MAXRESIDUE, maxResidue); @@ -311,6 +186,27 @@ public class StructureFrequency } if (bpEnd > 0) { + values[')'] = values['(']; + values['('] = 0; + + residueHash = new Hashtable(); + maxResidue = ")"; + + if (profile) + { + residueHash.put(PROFILE, new int[][] + { values, new int[] + { jSize, (jSize - values['-']) } }); + + residueHash.put(PAIRPROFILE, pairs); + } + + residueHash.put(MAXCOUNT, new Integer(count)); + residueHash.put(MAXRESIDUE, maxResidue); + + percentage = ((float) count * 100) / (float) jSize; + residueHash.put(PID_GAPS, new Float(percentage)); + result[bpEnd] = residueHash; } } @@ -404,19 +300,6 @@ public class StructureFrequency boolean ignoreGapsInConsensusCalculation, boolean includeAllConsSymbols) { - completeConsensus(consensus, hconsensus, iStart, width, - ignoreGapsInConsensusCalculation, includeAllConsSymbols, null); // new - // char[] - // { 'A', 'C', 'G', 'T', 'U' }); - } - - public static void completeConsensus(AlignmentAnnotation consensus, - Hashtable[] hconsensus, int iStart, int width, - boolean ignoreGapsInConsensusCalculation, - boolean includeAllConsSymbols, char[] alphabet) - { - System.out.println("StructureFrequency.completeConsensus " - + includeAllConsSymbols); float tval, value; if (consensus == null || consensus.annotations == null || consensus.annotations.length < width) @@ -457,52 +340,53 @@ public class StructureFrequency } int[][] profile = (int[][]) hconsensus[i] .get(StructureFrequency.PROFILE); - if (profile != null && includeAllConsSymbols) // Just responsible for the + int[][] pairs = (int[][]) hconsensus[i] + .get(StructureFrequency.PAIRPROFILE); + + if (pairs != null && includeAllConsSymbols) // Just responsible for the // tooltip + // TODO Update tooltips for Structure row { - // System.out.println("StructureFrequency.includeAllConsSymbols"); mouseOver = ""; - if (alphabet != null) + + /* TODO It's not sure what is the purpose of the alphabet and wheter it is useful for structure? + * + * if (alphabet != null) { for (int c = 0; c < alphabet.length; c++) { + * tval = ((float) profile[0][alphabet[c]]) 100f / (float) + * profile[1][ignoreGapsInConsensusCalculation ? 1 : 0]; mouseOver += + * ((c == 0) ? "" : "; ") + alphabet[c] + " " + ((int) tval) + "%"; } } + * else { + */ + Object[] ca = new Object[625]; + float[] vl = new float[625]; + int x = 0; + for (int c = 65; c < 90; c++) { - for (int c = 0; c < alphabet.length; c++) + for (int d = 65; d < 90; d++) { - tval = ((float) profile[0][alphabet[c]]) - * 100f - / (float) profile[1][ignoreGapsInConsensusCalculation ? 1 - : 0]; - mouseOver += ((c == 0) ? "" : "; ") + alphabet[c] + " " - + ((int) tval) + "%"; + ca[x] = new int[] + { c, d }; + vl[x] = (float) pairs[c][d]; + x++; } } - else + jalview.util.QuickSort.sort(vl, ca); + int p = 0; + + for (int c = 624; c > 0; c--) { - // System.out.println("StructureFrequency.NOTincludeAllConsSymbols"); - Object[] ca = new Object[profile[0].length]; - float[] vl = new float[profile[0].length]; - for (int c = 0; c < ca.length; c++) + if (vl[c] > 0) { - ca[c] = new char[] - { (char) c }; - vl[c] = (float) profile[0][c]; - } - ; - jalview.util.QuickSort.sort(vl, ca); - for (int p = 0, c = ca.length - 1; profile[0][((char[]) ca[c])[0]] > 0; c--) - { - if (((char[]) ca[c])[0] != '-') - { - tval = ((float) profile[0][((char[]) ca[c])[0]]) - * 100f - / (float) profile[1][ignoreGapsInConsensusCalculation ? 1 - : 0]; - mouseOver += ((p == 0) ? "" : "; ") + ((char[]) ca[c])[0] - + " " + ((int) tval) + "%"; - p++; + tval = ((float) vl[c] * 100f / (float) profile[1][ignoreGapsInConsensusCalculation ? 1 + : 0]); + mouseOver += ((p == 0) ? "" : "; ") + (char) ((int[]) ca[c])[0] + + (char) ((int[]) ca[c])[1] + " " + ((int) tval) + "%"; + p++; - } } - } + + // } } else { @@ -514,39 +398,49 @@ public class StructureFrequency } /** - * get the sorted profile for the given position of the consensus + * get the sorted base-pair profile for the given position of the consensus * * @param hconsensus - * @return + * @return profile of the given column */ public static int[] extractProfile(Hashtable hconsensus, - boolean ignoreGapsInConsensusCalculation, int column) + boolean ignoreGapsInConsensusCalculation) { - // TODO is there a more elegant way to acces the column number? - /* - * calculate the frequence of the 16 bp variations for this column 'somehow' - * transfer this via getProfile and let it correctly draw - */ - int[] rtnval = new int[51]; // 2*(5*5)+1 + int[] rtnval = new int[52]; // 2*(5*5)+2 int[][] profile = (int[][]) hconsensus.get(StructureFrequency.PROFILE); int[][] pairs = (int[][]) hconsensus .get(StructureFrequency.PAIRPROFILE); if (profile == null) return null; - - rtnval[0] = 1; - for (int j = 65; j <= 90; j++) + + // TODO fix the object length, also do it in completeConsensus + Object[] ca = new Object[625]; + float[] vl = new float[625]; + int x = 0; + for (int c = 65; c < 90; c++) { - for (int k = 65; k <= 90; k++) + for (int d = 65; d < 90; d++) { - if (pairs[j][k] > 0) - { - rtnval[rtnval[0]++] = j; - rtnval[rtnval[0]++] = k; - rtnval[rtnval[0]++] = (int) ((float) pairs[j][k] * 100f / (float) profile[1][ignoreGapsInConsensusCalculation ? 1 - : 0]); - } + ca[x] = new int[] + { c, d }; + vl[x] = (float) pairs[c][d]; + x++; + } + } + jalview.util.QuickSort.sort(vl, ca); + + rtnval[0] = 2; + rtnval[1] = 0; + for (int c = 624; c > 0; c--) + { + if (vl[c] > 0) + { + rtnval[rtnval[0]++] = ((int[]) ca[c])[0]; + rtnval[rtnval[0]++] = ((int[]) ca[c])[1]; + rtnval[rtnval[0]] = (int) ((float) vl[c] * 100f / (float) profile[1][ignoreGapsInConsensusCalculation ? 1 + : 0]); + rtnval[1]+=rtnval[rtnval[0]++]; } }