X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FAAFrequency.java;h=ded1eb65293ebaf3433de2533cc3f1c4e5d7ad17;hb=6e514532a128c47d099985375d6126111ef548cb;hp=46264d6eba335182d77258de4be3c2a8771828f7;hpb=cf4f0796ccc2d70bad2675e2feda184e5bef7a79;p=jalview.git
diff --git a/src/jalview/analysis/AAFrequency.java b/src/jalview/analysis/AAFrequency.java
index 46264d6..ded1eb6 100755
--- a/src/jalview/analysis/AAFrequency.java
+++ b/src/jalview/analysis/AAFrequency.java
@@ -1,25 +1,32 @@
/*
- * Jalview - A Sequence Alignment Editor and Viewer (Version 2.5)
- * 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.SequenceI;
+import jalview.util.Format;
-import jalview.datamodel.*;
+import java.util.Hashtable;
+import java.util.List;
/**
* Takes in a vector or array of sequences and column start and column end and
@@ -32,8 +39,8 @@ import jalview.datamodel.*;
*/
public class AAFrequency
{
- // No need to store 1000s of strings which are not
- // visible to the user.
+ private static final int TO_UPPER_CASE = 'A' - 'a'; // -32
+
public static final String MAXCOUNT = "C";
public static final String MAXRESIDUE = "R";
@@ -44,51 +51,61 @@ public class AAFrequency
public static final String PROFILE = "P";
- public static final Hashtable[] calculate(Vector sequences, int start,
- int end)
+ /*
+ * Quick look-up of String value of char 'A' to 'Z'
+ */
+ private static final String[] CHARS = new String['Z' - 'A' + 1];
+
+ static
{
- return calculate(sequences, start, end, false);
+ for (char c = 'A'; c <= 'Z'; c++)
+ {
+ CHARS[c - 'A'] = String.valueOf(c);
+ }
}
- public static final Hashtable[] calculate(Vector sequences, int start,
- int end, boolean profile)
+ public static final Hashtable[] calculate(List list,
+ int start, int end)
+ {
+ return calculate(list, start, end, false);
+ }
+
+ public static final Hashtable[] calculate(List sequences,
+ int start, int end, boolean profile)
{
SequenceI[] seqs = new SequenceI[sequences.size()];
int width = 0;
- for (int i = 0; i < sequences.size(); i++)
+ synchronized (sequences)
{
- seqs[i] = (SequenceI) sequences.elementAt(i);
- if (seqs[i].getLength() > width)
+ for (int i = 0; i < sequences.size(); i++)
{
- width = seqs[i].getLength();
+ seqs[i] = sequences.get(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);
+ Hashtable[] reply = new Hashtable[width];
- return reply;
- }
+ if (end >= width)
+ {
+ end = width;
+ }
- public static final void calculate(SequenceI[] sequences, int start,
- int end, Hashtable[] result)
- {
- calculate(sequences, start, end, result, false);
+ calculate(seqs, start, end, reply, profile);
+ return reply;
+ }
}
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;
+ int maxCount, nongap, i, j, v;
+ int jSize = sequences.length;
String maxResidue;
- char c;
+ char c = '-';
float percentage;
int[] values = new int[255];
@@ -105,6 +122,12 @@ public class AAFrequency
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)
{
@@ -122,7 +145,7 @@ public class AAFrequency
}
else if ('a' <= c && c <= 'z')
{
- c -= 32; // ('a' - 'A');
+ c += TO_UPPER_CASE;
}
nongap++;
@@ -134,25 +157,31 @@ public class AAFrequency
values['-']++;
}
}
-
- for (v = 'A'; v < 'Z'; v++)
+ if (jSize == 1)
{
- if (values[v] < 2 || values[v] < maxCount)
+ maxResidue = String.valueOf(c);
+ maxCount = 1;
+ }
+ else
+ {
+ for (v = 'A'; v <= 'Z'; v++)
{
- continue;
- }
+ 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);
+ if (values[v] > maxCount)
+ {
+ maxResidue = CHARS[v - 'A'];
+ }
+ else if (values[v] == maxCount)
+ {
+ maxResidue += CHARS[v - 'A'];
+ }
+ maxCount = values[v];
}
- maxCount = values[v];
}
-
if (maxResidue.length() == 0)
{
maxResidue = "-";
@@ -166,11 +195,16 @@ public class AAFrequency
residueHash.put(MAXCOUNT, new Integer(maxCount));
residueHash.put(MAXRESIDUE, maxResidue);
- percentage = ((float) maxCount * 100) / (float) jSize;
+ percentage = ((float) maxCount * 100) / jSize;
residueHash.put(PID_GAPS, new Float(percentage));
- percentage = ((float) maxCount * 100) / (float) nongap;
+ if (nongap > 0)
+ {
+ // calculate for non-gapped too
+ percentage = ((float) maxCount * 100) / nongap;
+ }
residueHash.put(PID_NOGAPS, new Float(percentage));
+
result[i] = residueHash;
}
}
@@ -186,51 +220,86 @@ public class AAFrequency
* @param width
* @param ignoreGapsInConsensusCalculation
* @param includeAllConsSymbols
+ * @param nseq
*/
public static void completeConsensus(AlignmentAnnotation consensus,
Hashtable[] hconsensus, int iStart, int width,
boolean ignoreGapsInConsensusCalculation,
- boolean includeAllConsSymbols)
+ boolean includeAllConsSymbols, long nseq)
{
completeConsensus(consensus, hconsensus, iStart, width,
- ignoreGapsInConsensusCalculation, includeAllConsSymbols, null); // new
- // char[]
- // { 'A', 'C', 'G', 'T', 'U' });
+ ignoreGapsInConsensusCalculation, includeAllConsSymbols, null,
+ nseq);
}
public static void completeConsensus(AlignmentAnnotation consensus,
Hashtable[] hconsensus, int iStart, int width,
boolean ignoreGapsInConsensusCalculation,
- boolean includeAllConsSymbols, char[] alphabet)
+ boolean includeAllConsSymbols, char[] alphabet, long nseq)
{
float tval, value;
- if (consensus==null || consensus.annotations==null || consensus.annotations.length= 10)
+ {
+ precision++;
+ nseq /= 10;
+ }
+ final Format fmt;
+ if (precision > 1)
+ {
+ // if (precision>2)
+ {
+ fmtstr = "%" + (2 + precision) + "." + (precision) + "f";
+ }
+ fmt = new Format(fmtstr);
+ }
+ else
+ {
+ fmt = null;
+ }
for (int i = iStart; i < width; i++)
{
+ Hashtable hci;
+ if (i >= hconsensus.length || ((hci = hconsensus[i]) == null))
+ {
+ // happens if sequences calculated over were shorter than alignment
+ // width
+ consensus.annotations[i] = null;
+ continue;
+ }
value = 0;
+ Float fv;
if (ignoreGapsInConsensusCalculation)
{
- value = ((Float) hconsensus[i].get(AAFrequency.PID_NOGAPS))
- .floatValue();
+ fv = (Float) hci.get(AAFrequency.PID_NOGAPS);
}
else
{
- value = ((Float) hconsensus[i].get(AAFrequency.PID_GAPS))
- .floatValue();
+ fv = (Float) hci.get(AAFrequency.PID_GAPS);
}
-
- String maxRes = hconsensus[i].get(AAFrequency.MAXRESIDUE).toString();
- String mouseOver = hconsensus[i].get(AAFrequency.MAXRESIDUE) + " ";
+ if (fv == null)
+ {
+ consensus.annotations[i] = null;
+ // data has changed below us .. give up and
+ continue;
+ }
+ value = fv.floatValue();
+ String maxRes = hci.get(AAFrequency.MAXRESIDUE).toString();
+ String mouseOver = hci.get(AAFrequency.MAXRESIDUE) + " ";
if (maxRes.length() > 1)
{
mouseOver = "[" + maxRes + "] ";
maxRes = "+";
}
- int[][] profile = (int[][]) hconsensus[i].get(AAFrequency.PROFILE);
+ int[][] profile = (int[][]) hci.get(AAFrequency.PROFILE);
if (profile != null && includeAllConsSymbols)
{
mouseOver = "";
@@ -238,12 +307,10 @@ public class AAFrequency
{
for (int c = 0; c < alphabet.length; c++)
{
- tval = ((float) profile[0][alphabet[c]])
- * 100f
- / (float) profile[1][ignoreGapsInConsensusCalculation ? 1
- : 0];
+ tval = profile[0][alphabet[c]] * 100f
+ / profile[1][ignoreGapsInConsensusCalculation ? 1 : 0];
mouseOver += ((c == 0) ? "" : "; ") + alphabet[c] + " "
- + ((int) tval) + "%";
+ + ((fmt != null) ? fmt.form(tval) : ((int) tval)) + "%";
}
}
else
@@ -254,7 +321,7 @@ public class AAFrequency
{
ca[c] = new char[]
{ (char) c };
- vl[c] = (float) profile[0][c];
+ vl[c] = profile[0][c];
}
;
jalview.util.QuickSort.sort(vl, ca);
@@ -262,12 +329,13 @@ public class AAFrequency
{
if (((char[]) ca[c])[0] != '-')
{
- tval = ((float) profile[0][((char[]) ca[c])[0]])
+ tval = profile[0][((char[]) ca[c])[0]]
* 100f
- / (float) profile[1][ignoreGapsInConsensusCalculation ? 1
- : 0];
+ / profile[1][ignoreGapsInConsensusCalculation ? 1 : 0];
mouseOver += ((p == 0) ? "" : "; ") + ((char[]) ca[c])[0]
- + " " + ((int) tval) + "%";
+ + " "
+ + ((fmt != null) ? fmt.form(tval) : ((int) tval))
+ + "%";
p++;
}
@@ -277,7 +345,8 @@ public class AAFrequency
}
else
{
- mouseOver += ((int) value + "%");
+ mouseOver += ((fmt != null) ? fmt.form(value) : ((int) value))
+ + "%";
}
consensus.annotations[i] = new Annotation(maxRes, mouseOver, ' ',
value);
@@ -296,25 +365,28 @@ public class AAFrequency
int[] rtnval = new int[64];
int[][] profile = (int[][]) hconsensus.get(AAFrequency.PROFILE);
if (profile == null)
+ {
return null;
- Object[] ca = new Object[profile[0].length];
+ }
+ char[][] ca = new char[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];
+ vl[c] = 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--)
+ rtnval[0] = 2;
+ rtnval[1] = 0;
+ for (int c = ca.length - 1; profile[0][ca[c][0]] > 0; c--)
{
- if (((char[]) ca[c])[0] != '-')
+ if (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
+ rtnval[rtnval[0]++] = ca[c][0];
+ rtnval[rtnval[0]] = (int) (profile[0][ca[c][0]] * 100f / profile[1][ignoreGapsInConsensusCalculation ? 1
: 0]);
+ rtnval[1] += rtnval[rtnval[0]++];
}
}
return rtnval;