--- /dev/null
+/*
+ * Jalview - A Sequence Alignment Editor and Viewer (Version 2.6)
+ * Copyright (C) 2010 J Procter, AM Waterhouse, G Barton, M Clamp, S Searle
+ *
+ * This file is part of Jalview.
+ *
+ * Jalview is free software: you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
+ *
+ * Jalview is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * PURPOSE. See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License along with Jalview. If not, see <http://www.gnu.org/licenses/>.
+ */
+package jalview.analysis;
+
+import java.util.*;
+
+import jalview.datamodel.*;
+
+/**
+ * Takes in a vector or array of sequences and column start and column end and
+ * returns a new Hashtable[] of size maxSeqLength, if Hashtable not supplied.
+ * This class is used extensively in calculating alignment colourschemes that
+ * depend on the amount of conservation in each alignment column.
+ *
+ * @author $author$
+ * @version $Revision$
+ */
+public class StructureFrequency
+{
+ // No need to store 1000s of strings which are not
+ // visible to the user.
+ public static final String MAXCOUNT = "C";
+
+ public static final String MAXRESIDUE = "R";
+
+ public static final String PID_GAPS = "G";
+
+ public static final String PID_NOGAPS = "N";
+
+ public static final String PROFILE = "P";
+
+ 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;
+ }
+ }
+
+ /**
+ * Method to calculate a 'base pair consensus row', very similar
+ * to nucleotide consensus but takes into account a given structure
+ * @param sequences
+ * @param start
+ * @param end
+ * @param result
+ * @param profile
+ * @param rnaStruc
+ */
+ public static final void calculate(SequenceI[] sequences, int start,
+ int end, Hashtable[] result, boolean profile, AlignmentAnnotation rnaStruc){
+ //TODO Consider to use AlignmentAnnotation instead of structure string
+
+ Hashtable residueHash;
+
+ char[] seq, struc=rnaStruc.getRNAStruc().toCharArray();
+ SequenceFeature[] rna =rnaStruc._rnasecstr;
+ char c,s,cEnd;
+ int count,nonGap,i,j,jSize = sequences.length;
+ int[] values = new int[255];
+ float percentage;
+
+
+ for (i = start; i < end; i++) //foreach column
+ {
+ residueHash = new Hashtable();
+ for (j = 0; j < jSize; j++) //foreach row
+ {
+ 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];
+ s = struc[i];
+ nonGap=0;
+
+ //standard representation for gaps in sequence and structure
+ if (c == '.' || c == ' ')
+ {
+ c = '-';
+ }
+ if (s == '.' || s == ' ')
+ {
+ s = '-';
+ }
+
+ if (c == '-')
+ {
+ values['-']++;
+ continue;
+ }
+ if(s == '-'){
+ values['-']++;
+ continue;
+ }
+ nonGap++;
+ cEnd=seq[rna[i].getEnd()];
+ if(checkBpType(c,cEnd)){
+ values['H']++; //H means it's a helix (structured)
+ }
+ }
+ }
+ /*UPDATE this for new values
+ if (profile)
+ {
+ residueHash.put(PROFILE, new int[][]
+ { values, new int[]
+ { jSize, nongap } });
+ }
+ */
+
+ count=values['H'];
+
+ percentage = ((float) count * 100) / (float) jSize;
+ residueHash.put(PID_GAPS, new Float(percentage));
+
+ //percentage = ((float) count * 100) / (float) nongap;
+ //residueHash.put(PID_NOGAPS, new Float(percentage));
+ result[i] = residueHash;
+
+ }
+ }
+
+
+ /**
+ * Method to check if a base-pair is a canonical or a wobble bp
+ * @param up 5' base
+ * @param down 3' base
+ * @return True if it is a canonical/wobble bp
+ */
+ public static boolean checkBpType(char up, char down){
+ if(up>'Z'){up-=32;}
+ if(down>'Z'){down-=32;}
+
+ switch (up){
+ case 'A':
+ switch (down){
+ case 'T': return true;
+ case 'U': return true;
+ }
+ break;
+ case 'C':
+ switch (down){
+ case 'G': return true;
+ }
+ break;
+ case 'T':
+ switch (down){
+ case 'A': return true;
+ case 'G': return true;
+ }
+ break;
+ case 'G':
+ switch (down){
+ case 'C': return true;
+ case 'T': return true;
+ case 'U': return true;
+ }
+ break;
+ case 'U':
+ switch (down){
+ case 'A': return true;
+ case 'G': return true;
+ }
+ break;
+ }
+ return false;
+ }
+
+ /**
+ * Compute all or part of the annotation row from the given consensus
+ * hashtable
+ *
+ * @param consensus
+ * - pre-allocated annotation row
+ * @param hconsensus
+ * @param iStart
+ * @param width
+ * @param ignoreGapsInConsensusCalculation
+ * @param includeAllConsSymbols
+ */
+ public static void completeConsensus(AlignmentAnnotation consensus,
+ Hashtable[] hconsensus, int iStart, int width,
+ 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)
+ {
+ float tval, value;
+ if (consensus == null || consensus.annotations == null
+ || consensus.annotations.length < width)
+ {
+ // called with a bad alignment annotation row - wait for it to be
+ // initialised properly
+ return;
+ }
+ for (int i = iStart; i < width; i++)
+ {
+ if (i >= hconsensus.length)
+ {
+ // happens if sequences calculated over were shorter than alignment
+ // width
+ consensus.annotations[i] = null;
+ continue;
+ }
+ value = 0;
+ if (ignoreGapsInConsensusCalculation)
+ {
+ value = ((Float) hconsensus[i].get(StructureFrequency.PID_NOGAPS))
+ .floatValue();
+ }
+ else
+ {
+ value = ((Float) hconsensus[i].get(StructureFrequency.PID_GAPS))
+ .floatValue();
+ }
+
+ String maxRes = hconsensus[i].get(StructureFrequency.MAXRESIDUE).toString();
+ String mouseOver = hconsensus[i].get(StructureFrequency.MAXRESIDUE) + " ";
+ if (maxRes.length() > 1)
+ {
+ mouseOver = "[" + maxRes + "] ";
+ maxRes = "+";
+ }
+ int[][] profile = (int[][]) hconsensus[i].get(StructureFrequency.PROFILE);
+ if (profile != null && includeAllConsSymbols)
+ {
+ mouseOver = "";
+ 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[profile[0].length];
+ float[] vl = new float[profile[0].length];
+ for (int c = 0; c < ca.length; c++)
+ {
+ 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++;
+
+ }
+ }
+
+ }
+ }
+ else
+ {
+ mouseOver += ((int) value + "%");
+ }
+ consensus.annotations[i] = new Annotation(maxRes, mouseOver, ' ',
+ value);
+ }
+ }
+
+ /**
+ * get the sorted profile for the given position of the consensus
+ *
+ * @param hconsensus
+ * @return
+ */
+ public static int[] extractProfile(Hashtable hconsensus,
+ boolean ignoreGapsInConsensusCalculation)
+ {
+ int[] rtnval = new int[64];
+ int[][] profile = (int[][]) hconsensus.get(StructureFrequency.PROFILE);
+ if (profile == null)
+ return null;
+ Object[] ca = new Object[profile[0].length];
+ float[] vl = new float[profile[0].length];
+ for (int c = 0; c < ca.length; c++)
+ {
+ ca[c] = new char[]
+ { (char) c };
+ vl[c] = (float) profile[0][c];
+ }
+ ;
+ jalview.util.QuickSort.sort(vl, ca);
+ rtnval[0] = 1;
+ for (int c = ca.length - 1; profile[0][((char[]) ca[c])[0]] > 0; c--)
+ {
+ if (((char[]) ca[c])[0] != '-')
+ {
+ rtnval[rtnval[0]++] = ((char[]) ca[c])[0];
+ rtnval[rtnval[0]++] = (int) (((float) profile[0][((char[]) ca[c])[0]]) * 100f / (float) profile[1][ignoreGapsInConsensusCalculation ? 1
+ : 0]);
+ }
+ }
+ return rtnval;
+ }
+
+ enum base {A,T,g,C};
+
+
+ public static void main(String args[]){
+ //Short test to see if checkBpType works
+ ArrayList<String> test = new ArrayList<String>();
+ 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+": "+StructureFrequency.checkBpType(i.charAt(0),j.charAt(0)));
+ }
+ }
+ }
+}
/** DOCUMENT ME!! */
public Hashtable[] hconsensus;
-
+
+ public Hashtable[] hStrucConsensus;
+
AlignmentAnnotation consensus;
+
+ AlignmentAnnotation strucConsensus;
AlignmentAnnotation conservation;
AlignmentAnnotation[] groupConservation;
boolean autoCalculateConsensus = true;
+
+ boolean autoCalculateStrucConsensus = true;
+
/** DOCUMENT ME!! */
public int ConsPercGaps = 25; // JBPNote : This should be a scalable property!
ConservationThread conservationThread;
ConsensusThread consensusThread;
+
+ StrucConsensusThread strucConsensusThread;
boolean consUpdateNeeded = false;
static boolean UPDATING_CONSENSUS = false;
+
+ static boolean UPDATING_STRUC_CONSENSUS = false;
static boolean UPDATING_CONSERVATION = false;
boolean updatingConsensus = false;
+
+ boolean updatingStrucConsensus = false;
boolean updatingConservation = false;
AAFrequency.calculate(alignment.getSequencesArray(), 0,
alignment.getWidth(), hconsensus, true);
updateAnnotation(true);
-
if (globalColourScheme != null)
{
globalColourScheme.setConsensus(hconsensus);
}
}
}
+
+ //--------START Structure Conservation
+ public void updateStrucConsensus(final AlignmentPanel ap)
+ {
+ // see note in mantis : issue number 8585
+ if (strucConsensus == null || !autoCalculateStrucConsensus)
+ {
+ return;
+ }
+ strucConsensusThread = new StrucConsensusThread(ap);
+ strucConsensusThread.start();
+ }
+
+ class StrucConsensusThread extends Thread
+ {
+ AlignmentPanel ap;
+
+ public StrucConsensusThread(AlignmentPanel ap)
+ {
+ this.ap = ap;
+ }
+
+ public void run()
+ {
+ updatingStrucConsensus = true;
+ while (UPDATING_STRUC_CONSENSUS)
+ {
+ try
+ {
+ if (ap != null)
+ {
+ ap.paintAlignment(false);
+ }
+
+ Thread.sleep(200);
+ } catch (Exception ex)
+ {
+ ex.printStackTrace();
+ }
+ }
+
+ UPDATING_STRUC_CONSENSUS = true;
+
+ try
+ {
+ int aWidth = (alignment != null) ? alignment.getWidth() : -1; // null
+ // pointer
+ // possibility
+ // here.
+ if (aWidth <= 0)
+ {
+ updatingStrucConsensus = false;
+ UPDATING_STRUC_CONSENSUS = false;
+ return;
+ }
+
+ strucConsensus.annotations = null;
+ strucConsensus.annotations = new Annotation[aWidth];
+
+ hStrucConsensus = new Hashtable[aWidth];
+ AlignmentAnnotation rna = ap.av.getAlignment().getAlignmentAnnotation()[0];
+ StructureFrequency.calculate(alignment.getSequencesArray(), 0,
+ alignment.getWidth(), hStrucConsensus, true, rna);
+ //TODO AlignmentAnnotation rnaStruc!!!
+ updateAnnotation(true);
+ if (globalColourScheme != null)
+ {
+ globalColourScheme.setConsensus(hStrucConsensus);
+ }
+
+ } catch (OutOfMemoryError error)
+ {
+ alignment.deleteAnnotation(consensus);
+
+ strucConsensus = null;
+ hStrucConsensus = null;
+ new OOMWarning("calculating structure consensus", error);
+ }
+ UPDATING_STRUC_CONSENSUS = false;
+ updatingStrucConsensus = false;
+
+ if (ap != null)
+ {
+ ap.paintAlignment(true);
+ }
+ }
+
+ /**
+ * update the consensus annotation from the sequence profile data using
+ * current visualization settings.
+ */
+ public void updateAnnotation()
+ {
+ updateAnnotation(false);
+ }
+
+ protected void updateAnnotation(boolean immediate)
+ {
+ // TODO: make calls thread-safe, so if another thread calls this method,
+ // it will either return or wait until one calculation is finished.
+ if (immediate
+ || (!updatingConsensus && consensus != null && hconsensus != null))
+ {
+ StructureFrequency.completeConsensus(strucConsensus, hStrucConsensus, 0,
+ hStrucConsensus.length, ignoreGapsInConsensusCalculation,
+ showSequenceLogo);
+ }
+ }
+ }
+ //--------END Structure Conservation
/**
* get the consensus sequence as displayed under the PID consensus annotation
{
updateConsensus(ap);
}
+
+ if(autoCalculateStrucConsensus)
+ {
+ updateStrucConsensus(ap);
+ }
// Reset endRes of groups if beyond alignment width
int alWidth = alignment.getWidth();
git://source.jalview.org / jalview.git / commitdiff