/*\r
* Jalview - A Sequence Alignment Editor and Viewer\r
-* Copyright (C) 2005 AM Waterhouse, J Procter, G Barton, M Clamp, S Searle\r
+* Copyright (C) 2006 AM Waterhouse, J Procter, G Barton, M Clamp, S Searle\r
*\r
* This program is free software; you can redistribute it and/or\r
* modify it under the terms of the GNU General Public License\r
\r
\r
/**\r
- * Takes in a vector of sequences and column start and column end\r
- * and returns a vector of size (end-start+1). Each element of the\r
- * vector contains a hashtable with the keys being residues and\r
- * the values being the count of each residue in that column.\r
+ * Takes in a vector or array of sequences and column start and column end\r
+ * and returns a new Hashtable[] of size maxSeqLength, if Hashtable not supplied.\r
* This class is used extensively in calculating alignment colourschemes\r
* that depend on the amount of conservation in each alignment column.\r
* @author $author$\r
*/\r
public class AAFrequency\r
{\r
- /** Takes in a vector of sequences and column start and column end\r
- * and returns a vector of size (end-start+1). Each element of the\r
- * vector contains a hashtable with the keys being residues and\r
- * the values being the count of each residue in that column.\r
- * This class is used extensively in calculating alignment colourschemes\r
- * that depend on the amount of conservation in each alignment column. */\r
- public static final Vector calculate(Vector sequences, int start, int end)\r
+ //No need to store 1000s of strings which are not\r
+ //visible to the user.\r
+ public static final String MAXCOUNT = "C";\r
+ public static final String MAXRESIDUE="R";\r
+ public static final String PID_GAPS = "G";\r
+ public static final String PID_NOGAPS="N";\r
+\r
+ public static final Hashtable [] calculate(Vector sequences, int start, int end)\r
+ {\r
+ SequenceI [] seqs = new SequenceI[sequences.size()];\r
+ int width = 0;\r
+ for(int i=0; i<sequences.size(); i++)\r
+ {\r
+ seqs[i] = (SequenceI) sequences.elementAt(i);\r
+ if(seqs[i].getLength()>width)\r
+ width = seqs[i].getLength();\r
+ }\r
+\r
+ Hashtable [] reply = new Hashtable[width];\r
+\r
+ calculate(seqs, start, end, reply);\r
+\r
+ return reply;\r
+ }\r
+\r
+public static final void calculate(SequenceI[] sequences,\r
+ int start, int end,\r
+ Hashtable [] result)\r
+{\r
+ Hashtable residueHash;\r
+ int maxCount, nongap, i, j, v, jSize = sequences.length;\r
+ String maxResidue;\r
+ char c;\r
+ float percentage;\r
+\r
+ int[] values = new int[132];\r
+\r
+ String seq;\r
+\r
+ for (i = start; i < end; i++)\r
+ {\r
+ residueHash = new Hashtable();\r
+ maxCount = 0;\r
+ maxResidue = "";\r
+ nongap = 0;\r
+ values = new int[132];\r
+\r
+ for (j = 0; j < jSize; j++)\r
{\r
- Vector result = new Vector();\r
- Hashtable residueHash;\r
- int count, maxCount, nongap, i, j, jSize = sequences.size();\r
- String maxResidue, sequence, res;\r
- float percentage;\r
+ seq = sequences[j].getSequence();\r
+ if (seq.length() > i)\r
+ {\r
+ c = seq.charAt(i);\r
\r
- for (i = start; i <= end; i++)\r
+ if(c == '.' || c==' ')\r
+ c = '-';\r
+\r
+ if(c=='-')\r
+ {\r
+ values['-']++;\r
+ continue;\r
+ }\r
+ else if ('a' <= c && c <= 'z')\r
{\r
- residueHash = new Hashtable();\r
- maxCount = 0;\r
- maxResidue = "-";\r
- nongap = 0;\r
-\r
- for (j = 0; j < jSize; j++)\r
- {\r
- if (sequences.elementAt(j) instanceof Sequence)\r
- {\r
- sequence = ((Sequence) sequences.elementAt(j)).getSequence();\r
-\r
- if (sequence.length() > i)\r
- {\r
- res = String.valueOf(Character.toUpperCase(sequence.charAt(i)));\r
-\r
- if (jalview.util.Comparison.isGap(res.charAt(0)))\r
- {\r
- res = "-"; // we always use this for gaps in the property vectors\r
- }\r
- else\r
- { nongap++; }\r
-\r
- if (residueHash.containsKey(res))\r
- {\r
- count = ((Integer) residueHash.get(res)).intValue();\r
- count++;\r
-\r
- if (!jalview.util.Comparison.isGap(res.charAt(0)) &&\r
- (count >= maxCount))\r
- {\r
- if (count > maxCount)\r
- {\r
- maxResidue = res;\r
- }\r
- else if (maxResidue.indexOf(res) == -1)\r
- {\r
- maxResidue += res;\r
- }\r
-\r
- maxCount = count;\r
- }\r
-\r
- residueHash.put(res, new Integer(count));\r
- }\r
- else\r
- {\r
- residueHash.put(res, new Integer(1));\r
- }\r
- }\r
- else\r
- {\r
- if (residueHash.containsKey("-"))\r
- {\r
- count = ((Integer) residueHash.get("-")).intValue();\r
- count++;\r
- residueHash.put("-", new Integer(count));\r
- }\r
- else\r
- {\r
- residueHash.put("-", new Integer(1));\r
- }\r
- }\r
- }\r
- }\r
-\r
- residueHash.put("maxCount", new Integer(maxCount));\r
- residueHash.put("maxResidue", maxResidue);\r
-\r
-\r
- //Size is redundant at present if we calculate percentage here\r
- //residueHash.put("size", new Integer(jSize));\r
- //residueHash.put("nogaps", new Integer(nongap));\r
-\r
- percentage = ((float)maxCount*100) / (float)jSize;\r
- residueHash.put("pid_gaps", new Float(percentage) );\r
-\r
- percentage = ((float)maxCount*100) / (float)nongap;\r
- residueHash.put("pid_nogaps", new Float(percentage) );\r
- result.addElement(residueHash);\r
+ c -= 32 ;//('a' - 'A');\r
}\r
\r
+ nongap++;\r
+ values[c]++;\r
\r
+ }\r
+ else\r
+ {\r
+ values['-']++;\r
+ }\r
+ }\r
\r
- return result;\r
+ for (v = 'A'; v < 'Z'; v++)\r
+ {\r
+ if (values[v] < 2 || values[v] < maxCount)\r
+ continue;\r
+\r
+ if (values[v] > maxCount)\r
+ {\r
+ maxResidue = String.valueOf( (char) v);\r
+ }\r
+ else if (values[v] == maxCount)\r
+ {\r
+ maxResidue += String.valueOf( (char) v);\r
+ }\r
+ maxCount = values[v];\r
}\r
+\r
+ if(maxResidue.length()==0)\r
+ maxResidue = "-";\r
+\r
+ residueHash.put(MAXCOUNT, new Integer(maxCount));\r
+ residueHash.put(MAXRESIDUE, maxResidue);\r
+\r
+ percentage = ( (float) maxCount * 100) / (float) jSize;\r
+ residueHash.put(PID_GAPS, new Float(percentage));\r
+\r
+ percentage = ( (float) maxCount * 100) / (float) nongap;\r
+ residueHash.put(PID_NOGAPS, new Float(percentage));\r
+ result[i] = residueHash;\r
+ }\r
+}\r
}\r
+\r
+\r