X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FStructureFrequency.java;h=5cf4e304e65211d29c086186922561d4627ea533;hb=34269ee3caf45d96ec5b4f99c973357459b04633;hp=4d83019dae2bbc159dab708a1fc22c3720cf86b0;hpb=46c34fb5b641a5ec57b5eec0e7bf964a9e72219f;p=jalview.git
diff --git a/src/jalview/analysis/StructureFrequency.java b/src/jalview/analysis/StructureFrequency.java
index 4d83019..5cf4e30 100644
--- a/src/jalview/analysis/StructureFrequency.java
+++ b/src/jalview/analysis/StructureFrequency.java
@@ -1,25 +1,33 @@
/*
- * 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 ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
*
* 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.
- *
+ * 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 .
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
*/
package jalview.analysis;
-import java.util.*;
+import jalview.datamodel.AlignmentAnnotation;
+import jalview.datamodel.Annotation;
+import jalview.datamodel.SequenceFeature;
+import jalview.datamodel.SequenceI;
+import jalview.util.Format;
-import jalview.datamodel.*;
+import java.util.ArrayList;
+import java.util.Hashtable;
/**
* Takes in a vector or array of sequences and column start and column end and
@@ -32,6 +40,8 @@ import jalview.datamodel.*;
*/
public class StructureFrequency
{
+ public static final int STRUCTURE_PROFILE_LENGTH = 74;
+
// No need to store 1000s of strings which are not
// visible to the user.
public static final String MAXCOUNT = "C";
@@ -46,150 +56,26 @@ 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++)
{
if (pairs[i].getBegin() == indice)
+
{
+
return pairs[i].getEnd();
+
}
}
return -1;
@@ -210,17 +96,18 @@ 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();
+ char[] struc = rnaStruc.getRNAStruc().toCharArray();
+
SequenceFeature[] rna = rnaStruc._rnasecstr;
char c, s, cEnd;
- int count, nonGap = 0, i, bpEnd = -1, j, jSize = sequences.length;
+ int count = 0, nonGap = 0, i, bpEnd = -1, j, jSize = sequences.length;
int[] values;
int[][] pairs;
float percentage;
-
+ boolean wooble = true;
for (i = start; i < end; i++) // foreach column
{
residueHash = new Hashtable();
@@ -228,14 +115,22 @@ public class StructureFrequency
values = new int[255];
pairs = new int[255][255];
bpEnd = -1;
+ // System.out.println("s="+struc[i]);
+ if (i < struc.length)
+ {
+ s = struc[i];
- s = struc[i];
+ }
+ else
+ {
+ s = '-';
+ }
if (s == '.' || s == ' ')
{
s = '-';
}
- if (s != '(')
+ if (s != '(' && s != '[')
{
if (s == '-')
{
@@ -244,19 +139,21 @@ public class StructureFrequency
}
else
{
- 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)
+ bpEnd = findPair(rna, i);
+
+ if (bpEnd > -1)
+ {
+ for (j = 0; j < jSize; j++) // foreach row
{
- c = seq[i];
+ if (sequences[j] == null)
+ {
+ System.err
+ .println("WARNING: Consensus skipping null sequence - possible race condition.");
+ continue;
+ }
+ c = sequences[j].getCharAt(i);
+ // System.out.println("c="+c);
// standard representation for gaps in sequence and structure
if (c == '.' || c == ' ')
@@ -269,17 +166,26 @@ public class StructureFrequency
values['-']++;
continue;
}
- bpEnd = findPair(rna, i);
- cEnd = seq[bpEnd];
- if (checkBpType(c, cEnd))
+ cEnd = sequences[j].getCharAt(bpEnd);
+
+ // System.out.println("pairs ="+c+","+cEnd);
+ if (checkBpType(c, cEnd) == true)
+ {
+ values['(']++; // H means it's a helix (structured)
+ maxResidue = "(";
+ wooble = true;
+ // System.out.println("It's a pair wc");
+
+ }
+ if (checkBpType(c, cEnd) == false)
{
- values['H']++; // H means it's a helix (structured)
+ wooble = false;
+ values['[']++; // H means it's a helix (structured)
+ maxResidue = "[";
+
}
- // System.out.println("pair: "+c+","+cEnd);
pairs[c][cEnd]++;
- // System.out.println("pairs: "+c+","+cEnd+" - "+pairs[c][cEnd]);
- maxResidue = "H";
}
}
// nonGap++;
@@ -287,20 +193,25 @@ public class StructureFrequency
// UPDATE this for new values
if (profile)
{
- // System.out.println("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(PAIRPROFILE, pairs);
}
-
- count = values['H'];
-
+ if (wooble == true)
+ {
+ count = values['('];
+ }
+ if (wooble == false)
+ {
+ count = values['['];
+ }
residueHash.put(MAXCOUNT, new Integer(count));
residueHash.put(MAXRESIDUE, maxResidue);
- percentage = ((float) count * 100) / (float) jSize;
+ percentage = ((float) count * 100) / jSize;
residueHash.put(PID_GAPS, new Float(percentage));
// percentage = ((float) count * 100) / (float) nongap;
@@ -311,7 +222,38 @@ public class StructureFrequency
}
if (bpEnd > 0)
{
+ values[')'] = values['('];
+ values[']'] = values['['];
+ values['('] = 0;
+ values['['] = 0;
+ residueHash = new Hashtable();
+ if (wooble == true)
+ {
+ // System.out.println(maxResidue+","+wooble);
+ maxResidue = ")";
+ }
+ if (wooble == false)
+ {
+ // System.out.println(maxResidue+","+wooble);
+ 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) / jSize;
+ residueHash.put(PID_GAPS, new Float(percentage));
+
result[bpEnd] = residueHash;
+
}
}
}
@@ -402,21 +344,8 @@ public class StructureFrequency
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)
+ boolean includeAllConsSymbols, long nseq)
{
- System.out.println("StructureFrequency.completeConsensus "
- + includeAllConsSymbols);
float tval, value;
if (consensus == null || consensus.annotations == null
|| consensus.annotations.length < width)
@@ -425,9 +354,23 @@ public class StructureFrequency
// initialised properly
return;
}
+ String fmtstr = "%3.1f";
+ int precision = 2;
+ while (nseq > 100)
+ {
+ precision++;
+ nseq /= 10;
+ }
+ if (precision > 2)
+ {
+ fmtstr = "%" + (2 + precision) + "." + precision + "f";
+ }
+ Format fmt = new Format(fmtstr);
+
for (int i = iStart; i < width; i++)
{
- if (i >= hconsensus.length)
+ Hashtable hci;
+ if (i >= hconsensus.length || ((hci = hconsensus[i]) == null))
{
// happens if sequences calculated over were shorter than alignment
// width
@@ -435,78 +378,83 @@ public class StructureFrequency
continue;
}
value = 0;
+ Float fv;
if (ignoreGapsInConsensusCalculation)
{
- value = ((Float) hconsensus[i].get(StructureFrequency.PID_NOGAPS))
- .floatValue();
+ fv = (Float) hci.get(StructureFrequency.PID_NOGAPS);
}
else
{
- value = ((Float) hconsensus[i].get(StructureFrequency.PID_GAPS))
- .floatValue();
+ fv = (Float) hci.get(StructureFrequency.PID_GAPS);
}
-
- String maxRes = hconsensus[i].get(StructureFrequency.MAXRESIDUE)
- .toString();
- String mouseOver = hconsensus[i].get(StructureFrequency.MAXRESIDUE)
- + " ";
+ if (fv == null)
+ {
+ consensus.annotations[i] = null;
+ // data has changed below us .. give up and
+ continue;
+ }
+ value = fv.floatValue();
+ String maxRes = hci.get(StructureFrequency.MAXRESIDUE).toString();
+ String mouseOver = hci.get(StructureFrequency.MAXRESIDUE) + " ";
if (maxRes.length() > 1)
{
mouseOver = "[" + maxRes + "] ";
maxRes = "+";
}
- int[][] profile = (int[][]) hconsensus[i]
- .get(StructureFrequency.PROFILE);
- if (profile != null && includeAllConsSymbols) // Just responsible for the
+ int[][] profile = (int[][]) hci.get(StructureFrequency.PROFILE);
+ int[][] pairs = (int[][]) hci.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] = 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++)
- {
- 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 (vl[c] > 0)
{
- 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 = (vl[c] * 100f / profile[1][ignoreGapsInConsensusCalculation ? 1
+ : 0]);
+ mouseOver += ((p == 0) ? "" : "; ") + (char) ((int[]) ca[c])[0]
+ + (char) ((int[]) ca[c])[1] + " " + fmt.form(tval)
+ + "%";
+ p++;
- }
}
-
}
+
+ // }
}
else
{
- mouseOver += ((int) value + "%");
+ mouseOver += (fmt.form(value) + "%");
}
consensus.annotations[i] = new Annotation(maxRes, mouseOver, ' ',
value);
@@ -514,43 +462,63 @@ 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[STRUCTURE_PROFILE_LENGTH]; // 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];
+ int[][] ca = new int[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] = pairs[c][d];
+ x++;
+ }
+ }
+ jalview.util.QuickSort.sort(vl, ca);
+
+ int valuesCount = 0;
+ rtnval[1] = 0;
+ int offset = 2;
+ for (int c = 624; c > 0; c--)
+ {
+ if (vl[c] > 0)
+ {
+ rtnval[offset++] = ca[c][0];
+ rtnval[offset++] = ca[c][1];
+ rtnval[offset] = (int) (vl[c] * 100f / profile[1][ignoreGapsInConsensusCalculation ? 1
+ : 0]);
+ rtnval[1] += rtnval[offset++];
+ valuesCount++;
}
}
+ rtnval[0] = valuesCount;
- return rtnval;
+ // insert profile type code in position 0
+ int[] result = new int[rtnval.length + 1];
+ result[0] = AlignmentAnnotation.STRUCTURE_PROFILE;
+ System.arraycopy(rtnval, 0, result, 1, rtnval.length);
+ return result;
}
public static void main(String args[])