X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fanalysis%2FStructureFrequency.java;h=6816547f78b02a87953deba1dbcd517c21df505b;hb=9e00a71bcb54f0bcfcd082ae1660a1c9fdf9d4ed;hp=c253dceabb3b10a539624b634f500829eafa7a5a;hpb=1cf86680f3af423e29529855da07896501ce9f91;p=jalview.git
diff --git a/src/jalview/analysis/StructureFrequency.java b/src/jalview/analysis/StructureFrequency.java
index c253dce..6816547 100644
--- a/src/jalview/analysis/StructureFrequency.java
+++ b/src/jalview/analysis/StructureFrequency.java
@@ -1,24 +1,27 @@
/*
- * 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 2.8.0b1)
+ * Copyright (C) 2014 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.
- *
+ *
* 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 .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
*/
+
package jalview.analysis;
import java.util.*;
+import jalview.util.Format;
import jalview.datamodel.*;
/**
@@ -46,19 +49,26 @@ public class StructureFrequency
public static final String PAIRPROFILE = "B";
-/**
- * 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
- */
+ /**
+ * 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;
@@ -79,16 +89,18 @@ public class StructureFrequency
int end, Hashtable[] result, boolean profile,
AlignmentAnnotation rnaStruc)
{
+
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();
@@ -96,9 +108,14 @@ public class StructureFrequency
values = new int[255];
pairs = new int[255][255];
bpEnd = -1;
- if(i 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 == ' ')
@@ -140,15 +159,26 @@ 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 = "[";
+
}
pairs[c][cEnd]++;
- maxResidue = "H";
}
}
// nonGap++;
@@ -162,13 +192,18 @@ public class StructureFrequency
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;
@@ -179,7 +214,38 @@ 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;
+
}
}
}
@@ -270,18 +336,7 @@ 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)
{
float tval, value;
if (consensus == null || consensus.annotations == null
@@ -291,9 +346,22 @@ 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
@@ -301,77 +369,82 @@ 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
+ // TODO Update tooltips for Structure row
{
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--)
{
- 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 = (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);
@@ -385,47 +458,46 @@ public class StructureFrequency
* @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[74]; // 2*(5*5)+2
int[][] profile = (int[][]) hconsensus.get(StructureFrequency.PROFILE);
int[][] pairs = (int[][]) hconsensus
.get(StructureFrequency.PAIRPROFILE);
if (profile == null)
return null;
-
+
+ // TODO fix the object length, also do it in completeConsensus
Object[] ca = new Object[625];
float[] vl = new float[625];
- int x=0;
+ int x = 0;
for (int c = 65; c < 90; c++)
{
- for(int d = 65; d< 90; d++)
+ for (int d = 65; d < 90; d++)
{
- ca[x] = new int[]{ c, d};
- vl[x] = (float) pairs[c][d];
- x++;
+ ca[x] = new int[]
+ { c, d };
+ vl[x] = pairs[c][d];
+ x++;
}
}
jalview.util.QuickSort.sort(vl, ca);
-
- rtnval[0] = 1;
- for (int c=624; c>0; c--)
+
+ rtnval[0] = 2;
+ rtnval[1] = 0;
+ for (int c = 624; c > 0; c--)
{
- if (vl[c]>0)
+ 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
+ rtnval[rtnval[0]] = (int) (vl[c] * 100f / profile[1][ignoreGapsInConsensusCalculation ? 1
: 0]);
- }
+ rtnval[1] += rtnval[rtnval[0]++];
+ }
}
-
+
return rtnval;
}