-/*\r
-* Jalview - A Sequence Alignment Editor and Viewer\r
-* Copyright (C) 2005 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
-* as published by the Free Software Foundation; either version 2\r
-* of the License, or (at your option) any later version.\r
-*\r
-* This program is distributed in the hope that it will be useful,\r
-* but WITHOUT ANY WARRANTY; without even the implied warranty of\r
-* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the\r
-* GNU General Public License for more details.\r
-*\r
-* You should have received a copy of the GNU General Public License\r
-* along with this program; if not, write to the Free Software\r
-* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA\r
-*/\r
-package jalview.analysis;\r
-\r
-import jalview.datamodel.*;\r
-\r
-import java.util.*;\r
-\r
-\r
-/**\r
- * Calculates conservation values for a given set of sequences\r
- *\r
- * @author $author$\r
- * @version $Revision$\r
- */\r
-public class Conservation\r
-{\r
- Vector sequences;\r
- int start;\r
- int end;\r
- Vector seqNums; // vector of int vectors where first is sequence checksum\r
- int maxLength = 0; // used by quality calcs\r
- boolean seqNumsChanged = false; // updated after any change via calcSeqNum;\r
- Vector total = new Vector();\r
-\r
- /** Stores calculated quality values */\r
- public Vector quality;\r
-\r
- /** Stores maximum and minimum values of quality values */\r
- public Double[] qualityRange = new Double[2];\r
- String consString = "";\r
- Sequence consSequence;\r
- Hashtable propHash;\r
- int threshold;\r
- String name = "";\r
- int[][] cons2;\r
-\r
- /**\r
- * Creates a new Conservation object.\r
- *\r
- * @param name Name of conservation\r
- * @param propHash DOCUMENT ME!\r
- * @param threshold to count the residues in residueHash(). commonly used value is 3\r
- * @param sequences sequences to be used in calculation\r
- * @param start start residue position\r
- * @param end end residue position\r
- */\r
- public Conservation(String name, Hashtable propHash, int threshold,\r
- Vector sequences, int start, int end)\r
- {\r
- this.name = name;\r
- this.propHash = propHash;\r
- this.threshold = threshold;\r
- this.sequences = sequences;\r
- this.start = start;\r
- this.end = end;\r
- seqNums = new Vector(sequences.size());\r
- calcSeqNums();\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- */\r
- private void calcSeqNums()\r
- {\r
- int i=0, iSize=sequences.size();\r
- for (i=0; i < iSize; i++)\r
- {\r
- calcSeqNum(i);\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param i DOCUMENT ME!\r
- */\r
- private void calcSeqNum(int i)\r
- {\r
- String sq = null; // for dumb jbuilder not-inited exception warning\r
- int[] sqnum = null;\r
-\r
- if ((i > -1) && (i < sequences.size()))\r
- {\r
- sq = ((SequenceI) sequences.elementAt(i)).getSequence();\r
-\r
- if (seqNums.size() <= i)\r
- {\r
- seqNums.addElement(new int[sq.length() + 1]);\r
- }\r
-\r
- if (sq.hashCode() != ((int[]) seqNums.elementAt(i))[0])\r
- {\r
- int j;\r
- int len;\r
- seqNumsChanged = true;\r
- sq = ((SequenceI) sequences.elementAt(i)).getSequence();\r
- len = sq.length();\r
-\r
- if (maxLength < len)\r
- {\r
- maxLength = len;\r
- }\r
-\r
- sqnum = new int[len + 1]; // better to always make a new array - sequence can change its length\r
- sqnum[0] = sq.hashCode();\r
-\r
- for (j = 1; j <= len; j++)\r
- {\r
- sqnum[j] = ((Integer) jalview.schemes.ResidueProperties.aaHash.get(String.valueOf(\r
- sq.charAt(j - 1)))).intValue(); // yuk - JBPNote - case taken care of in aaHash\r
- }\r
-\r
- seqNums.setElementAt(sqnum, i);\r
- }\r
- }\r
- else\r
- {\r
- // JBPNote INFO level debug\r
- System.err.println(\r
- "ERROR: calcSeqNum called with out of range sequence index for Alignment\n");\r
- }\r
- }\r
-\r
- /**\r
- * Calculates the conservation values for given set of sequences\r
- */\r
- public void calculate()\r
- {\r
- Hashtable resultHash, residueHash, ht;\r
- int count, thresh, j, jSize = sequences.size();\r
- String type, res=null;\r
- SequenceI sequence;\r
- char c;\r
- Enumeration enumeration, enumeration2;\r
-\r
- for (int i = start; i <= end; i++)\r
- {\r
- resultHash = new Hashtable();\r
- residueHash = new Hashtable();\r
-\r
- for (j = 0; j < jSize; j++)\r
- {\r
- // JBPNote - have to make sure elements of the sequences vector\r
- // are tested like this everywhere...\r
- sequence = (Sequence) sequences.elementAt(j);\r
-\r
- if (sequence.getLength() > i)\r
- {\r
- c = sequence.getCharAt(i);\r
-\r
- if(jalview.util.Comparison.isGap(c))\r
- c = '-';\r
-\r
- if ('a' <= c && c <= 'z')\r
- {\r
- // TO UPPERCASE !!!\r
- //Faster than toUpperCase\r
- c -= ('a' - 'A') ;\r
- }\r
-\r
- res = String.valueOf( c );\r
-\r
-\r
- if (residueHash.containsKey(res))\r
- {\r
- count = ((Integer) residueHash.get(res)).intValue();\r
- count++;\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
- //What is the count threshold to count the residues in residueHash()\r
- thresh = (threshold * (sequences.size())) / 100;\r
-\r
- //loop over all the found residues\r
- enumeration = residueHash.keys();\r
-\r
- while (enumeration.hasMoreElements())\r
- {\r
- res = (String) enumeration.nextElement();\r
-\r
- if (((Integer) residueHash.get(res)).intValue() > thresh)\r
- {\r
- //Now loop over the properties\r
- enumeration2 = propHash.keys();\r
-\r
- while (enumeration2.hasMoreElements())\r
- {\r
- type = (String) enumeration2.nextElement();\r
- ht = (Hashtable) propHash.get(type);\r
-\r
- //Have we ticked this before?\r
- if (!resultHash.containsKey(type))\r
- {\r
- if (ht.containsKey(res))\r
- {\r
- resultHash.put(type, ht.get(res));\r
- }\r
- else\r
- {\r
- resultHash.put(type, ht.get("-"));\r
- }\r
- }\r
- else if (((Integer) resultHash.get(type)).equals(\r
- (Integer) ht.get(res)) == false)\r
- {\r
- resultHash.put(type, new Integer(-1));\r
- }\r
- }\r
- }\r
- }\r
-\r
- total.addElement(resultHash);\r
- }\r
- }\r
-\r
-\r
- /***\r
- * countConsNGaps\r
- * returns gap count in int[0], and conserved residue count in int[1]\r
- */\r
- public int[] countConsNGaps(int j)\r
- {\r
- int count = 0;\r
- int cons = 0;\r
- int nres = 0;\r
- int[] r = new int[2];\r
- char f = '$';\r
- int i, iSize = sequences.size();\r
- char c;\r
-\r
- for (i = 0; i < iSize; i++)\r
- {\r
- if (j >= ((Sequence) sequences.elementAt(i)).getLength())\r
- {\r
- count++;\r
- continue;\r
- }\r
-\r
- c = ((Sequence) sequences.elementAt(i)).getCharAt(j); // gaps do not have upper/lower case\r
-\r
- if (jalview.util.Comparison.isGap((c)))\r
- {\r
- count++;\r
- }\r
- else\r
- {\r
- nres++;\r
-\r
- if (nres == 1)\r
- {\r
- f = c;\r
- cons++;\r
- }\r
- else if (f == c)\r
- {\r
- cons++;\r
- }\r
- }\r
- }\r
-\r
- r[0] = (nres == cons) ? 1 : 0;\r
- r[1] = count;\r
-\r
- return r;\r
- }\r
-\r
- /**\r
- * Calculates the conservation sequence\r
- *\r
- * @param consflag if true, poitiveve conservation; false calculates negative conservation\r
- * @param percentageGaps commonly used value is 25\r
- */\r
- public void verdict(boolean consflag, float percentageGaps)\r
- {\r
- StringBuffer consString = new StringBuffer();\r
- String type;\r
- Integer result;\r
- int[] gapcons;\r
- int totGaps, count;\r
- float pgaps;\r
- Hashtable resultHash ;\r
- Enumeration enumeration;\r
-\r
-\r
- for (int i = start; i <= end; i++)\r
- {\r
- gapcons = countConsNGaps(i);\r
- totGaps = gapcons[1];\r
- pgaps = ((float) totGaps * 100) / (float) sequences.size();\r
-\r
- if (percentageGaps > pgaps)\r
- {\r
- resultHash = (Hashtable) total.elementAt(i - start);\r
-\r
- //Now find the verdict\r
- count = 0;\r
- enumeration = resultHash.keys();\r
-\r
- while (enumeration.hasMoreElements())\r
- {\r
- type = (String) enumeration.nextElement();\r
- result = (Integer) resultHash.get(type);\r
-\r
- //Do we want to count +ve conservation or +ve and -ve cons.?\r
- if (consflag)\r
- {\r
- if (result.intValue() == 1)\r
- {\r
- count++;\r
- }\r
- }\r
- else\r
- {\r
- if (result.intValue() != -1)\r
- {\r
- count++;\r
- }\r
- }\r
- }\r
-\r
- if (count < 10)\r
- {\r
- consString.append(count); // Conserved props!=Identity\r
- }\r
- else\r
- {\r
- consString.append((gapcons[0] == 1) ? "*" : "+");\r
- }\r
- }\r
- else\r
- {\r
- consString.append("-");\r
- }\r
- }\r
-\r
- consSequence = new Sequence(name, consString.toString(), start, end);\r
- }\r
-\r
- /**\r
- *\r
- *\r
- * @return Conservation sequence\r
- */\r
- public Sequence getConsSequence()\r
- {\r
- return consSequence;\r
- }\r
-\r
- // From Alignment.java in jalview118\r
- public void findQuality()\r
- {\r
- findQuality(0, maxLength - 1);\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- */\r
- private void percentIdentity2()\r
- {\r
- calcSeqNums(); // updates maxLength, too.\r
-\r
- if ((cons2 == null) || seqNumsChanged)\r
- {\r
- cons2 = new int[maxLength][24];\r
-\r
- // Initialize the array\r
- for (int j = 0; j < 24; j++)\r
- {\r
- for (int i = 0; i < maxLength; i++)\r
- {\r
- cons2[i][j] = 0;\r
- }\r
- }\r
-\r
- int[] sqnum;\r
- int j = 0;\r
-\r
- while (j < sequences.size())\r
- {\r
- sqnum = (int[]) seqNums.elementAt(j);\r
-\r
- for (int i = 1; i < sqnum.length; i++)\r
- {\r
- cons2[i - 1][sqnum[i]]++;\r
- }\r
-\r
- for (int i = sqnum.length - 1; i < maxLength; i++)\r
- {\r
- cons2[i][23]++; // gap count\r
- }\r
-\r
- j++;\r
- }\r
-\r
- // unnecessary ?\r
-\r
- /* for (int i=start; i <= end; i++) {\r
- int max = -1000;\r
- int maxi = -1;\r
- int maxj = -1;\r
-\r
- for (int j=0;j<24;j++) {\r
- if (cons2[i][j] > max) {\r
- max = cons2[i][j];\r
- maxi = i;\r
- maxj = j;\r
- }\r
-\r
- }\r
- } */\r
- }\r
- }\r
-\r
- /**\r
- * Calculates the quality of the set of sequences\r
- *\r
- * @param start Start residue\r
- * @param end End residue\r
- */\r
- public void findQuality(int start, int end)\r
- {\r
- quality = new Vector();\r
-\r
- double max = -10000;\r
- int[][] BLOSUM62 = jalview.schemes.ResidueProperties.getBLOSUM62();\r
-\r
- //Loop over columns // JBPNote Profiling info\r
- // long ts = System.currentTimeMillis();\r
- //long te = System.currentTimeMillis();\r
- percentIdentity2();\r
-\r
- int size = seqNums.size();\r
- int[] lengths = new int[size];\r
- double tot, bigtot, sr, tmp;\r
- double [] x, xx;\r
- int l, j, i, ii, seqNum;\r
-\r
- for (l = 0; l < size; l++)\r
- lengths[l] = ((int[]) seqNums.elementAt(l)).length - 1;\r
-\r
-\r
- for (j = start; j <= end; j++)\r
- {\r
- bigtot = 0;\r
-\r
- // First Xr = depends on column only\r
- x = new double[24];\r
-\r
- for (ii = 0; ii < 24; ii++)\r
- {\r
- x[ii] = 0;\r
-\r
- try\r
- {\r
- for (int i2 = 0; i2 < 24; i2++)\r
- {\r
- x[ii] += (((double) cons2[j][i2] * BLOSUM62[ii][i2]) +\r
- 4);\r
- }\r
- }\r
- catch (Exception e)\r
- {\r
- System.err.println("Exception during quality calculation.");\r
- e.printStackTrace();\r
- }\r
-\r
- //System.out.println("X " + ii + " " + x[ii]);\r
- x[ii] /= (size);\r
-\r
- //System.out.println("X " + ii + " " + x[ii]);\r
- }\r
-\r
- // Now calculate D for each position and sum\r
- for (int k = 0; k < size; k++)\r
- {\r
- tot = 0;\r
- xx = new double[24];\r
- seqNum = (j < lengths[k])\r
- ? ((int[]) seqNums.elementAt(k))[j + 1] : 23; // Sequence, or gap at the end\r
-\r
- // This is a loop over r\r
- for (i = 0; i < 23; i++)\r
- {\r
- sr = 0;\r
-\r
- try\r
- {\r
- sr = (double) BLOSUM62[i][seqNum] + 4;\r
- }\r
- catch (Exception e)\r
- {\r
- System.out.println("Exception in sr: " + e);\r
- e.printStackTrace();\r
- }\r
-\r
- //Calculate X with another loop over residues\r
- // System.out.println("Xi " + i + " " + x[i] + " " + sr);\r
- xx[i] = x[i] - sr;\r
-\r
- tot += (xx[i] * xx[i]);\r
- }\r
-\r
- bigtot += Math.sqrt(tot);\r
- }\r
-\r
- // This is the quality for one column\r
- if (max < bigtot)\r
- {\r
- max = bigtot;\r
- }\r
-\r
- // bigtot = bigtot * (size-cons2[j][23])/size;\r
- quality.addElement(new Double(bigtot));\r
-\r
-\r
- // Need to normalize by gaps\r
- }\r
-\r
- double newmax = -10000;\r
-\r
- for (j = start; j <= end; j++)\r
- {\r
- tmp = ((Double) quality.elementAt(j)).doubleValue();\r
- tmp = ((max - tmp) * (size - cons2[j][23])) / size;\r
-\r
- // System.out.println(tmp+ " " + j);\r
- quality.setElementAt(new Double(tmp), j);\r
-\r
- if (tmp > newmax)\r
- {\r
- newmax = tmp;\r
- }\r
- }\r
-\r
- // System.out.println("Quality " + s);\r
- qualityRange[0] = new Double(0);\r
- qualityRange[1] = new Double(newmax);\r
- }\r
-}\r
+/*
+ * 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.
+ *
+ * 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/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
+package jalview.analysis;
+
+import jalview.analysis.scoremodels.ScoreMatrix;
+import jalview.analysis.scoremodels.ScoreModels;
+import jalview.datamodel.AlignmentAnnotation;
+import jalview.datamodel.Annotation;
+import jalview.datamodel.ResidueCount;
+import jalview.datamodel.ResidueCount.SymbolCounts;
+import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceI;
+import jalview.schemes.ResidueProperties;
+import jalview.util.Comparison;
+
+import java.awt.Color;
+import java.util.List;
+import java.util.Map;
+import java.util.Map.Entry;
+import java.util.SortedMap;
+import java.util.TreeMap;
+import java.util.Vector;
+
+/**
+ * Calculates conservation values for a given set of sequences
+ */
+public class Conservation
+{
+ /*
+ * need to have a minimum of 3% of sequences with a residue
+ * for it to be included in the conservation calculation
+ */
+ private static final int THRESHOLD_PERCENT = 3;
+
+ private static final int TOUPPERCASE = 'a' - 'A';
+
+ private static final int GAP_INDEX = -1;
+
+ SequenceI[] sequences;
+
+ int start;
+
+ int end;
+
+ /*
+ * a list whose i'th element is an array whose first entry is the checksum
+ * of the i'th sequence, followed by residues encoded to score matrix index
+ */
+ Vector<int[]> seqNums;
+
+ int maxLength = 0; // used by quality calcs
+
+ boolean seqNumsChanged = false; // updated after any change via calcSeqNum;
+
+ /*
+ * a map per column with {property, conservation} where conservation value is
+ * 1 (property is conserved), 0 (absence of property is conserved) or -1
+ * (property is not conserved i.e. column has residues with and without it)
+ */
+ Map<String, Integer>[] total;
+
+ /*
+ * if true then conservation calculation will map all symbols to canonical aa
+ * numbering rather than consider conservation of that symbol
+ */
+ boolean canonicaliseAa = true;
+
+ private Vector<Double> quality;
+
+ private double qualityMinimum;
+
+ private double qualityMaximum;
+
+ private Sequence consSequence;
+
+ /*
+ * percentage of residues in a column to qualify for counting conservation
+ */
+ private int threshold;
+
+ private String name = "";
+
+ /*
+ * an array, for each column, of counts of symbols (by score matrix index)
+ */
+ private int[][] cons2;
+
+ /*
+ * gap counts for each column
+ */
+ private int[] cons2GapCounts;
+
+ private String[] consSymbs;
+
+ /**
+ * Constructor using default threshold of 3%
+ *
+ * @param name
+ * Name of conservation
+ * @param sequences
+ * sequences to be used in calculation
+ * @param start
+ * start residue position
+ * @param end
+ * end residue position
+ */
+ public Conservation(String name, List<SequenceI> sequences, int start,
+ int end)
+ {
+ this(name, THRESHOLD_PERCENT, sequences, start, end);
+ }
+
+ /**
+ * Constructor
+ *
+ * @param name
+ * Name of conservation
+ * @param threshold
+ * percentage of sequences at or below which property conservation is
+ * ignored
+ * @param sequences
+ * sequences to be used in calculation
+ * @param start
+ * start column position
+ * @param end
+ * end column position
+ */
+ public Conservation(String name, int threshold,
+ List<SequenceI> sequences, int start, int end)
+ {
+ this.name = name;
+ this.threshold = threshold;
+ this.start = start;
+ this.end = end;
+
+ maxLength = end - start + 1; // default width includes bounds of
+ // calculation
+
+ int s, sSize = sequences.size();
+ SequenceI[] sarray = new SequenceI[sSize];
+ this.sequences = sarray;
+ try
+ {
+ for (s = 0; s < sSize; s++)
+ {
+ sarray[s] = sequences.get(s);
+ if (sarray[s].getLength() > maxLength)
+ {
+ maxLength = sarray[s].getLength();
+ }
+ }
+ } catch (ArrayIndexOutOfBoundsException ex)
+ {
+ // bail - another thread has modified the sequence array, so the current
+ // calculation is probably invalid.
+ this.sequences = new SequenceI[0];
+ maxLength = 0;
+ }
+ }
+
+ /**
+ * Translate sequence i into score matrix indices and store it in the i'th
+ * position of the seqNums array.
+ *
+ * @param i
+ * @param sm
+ */
+ private void calcSeqNum(int i, ScoreMatrix sm)
+ {
+ int sSize = sequences.length;
+
+ if ((i > -1) && (i < sSize))
+ {
+ String sq = sequences[i].getSequenceAsString();
+
+ if (seqNums.size() <= i)
+ {
+ seqNums.addElement(new int[sq.length() + 1]);
+ }
+
+ /*
+ * the first entry in the array is the sequence's hashcode,
+ * following entries are matrix indices of sequence characters
+ */
+ if (sq.hashCode() != seqNums.elementAt(i)[0])
+ {
+ int j;
+ int len;
+ seqNumsChanged = true;
+ len = sq.length();
+
+ if (maxLength < len)
+ {
+ maxLength = len;
+ }
+
+ int[] sqnum = new int[len + 1]; // better to always make a new array -
+ // sequence can change its length
+ sqnum[0] = sq.hashCode();
+
+ for (j = 1; j <= len; j++)
+ {
+ // sqnum[j] = ResidueProperties.aaIndex[sq.charAt(j - 1)];
+ char residue = sq.charAt(j - 1);
+ if (Comparison.isGap(residue))
+ {
+ sqnum[j] = GAP_INDEX;
+ }
+ else
+ {
+ sqnum[j] = sm.getMatrixIndex(residue);
+ if (sqnum[j] == -1)
+ {
+ sqnum[j] = GAP_INDEX;
+ }
+ }
+ }
+
+ seqNums.setElementAt(sqnum, i);
+ }
+ else
+ {
+ System.out.println("SEQUENCE HAS BEEN DELETED!!!");
+ }
+ }
+ else
+ {
+ // JBPNote INFO level debug
+ System.err
+ .println("ERROR: calcSeqNum called with out of range sequence index for Alignment\n");
+ }
+ }
+
+ /**
+ * Calculates the conservation values for given set of sequences
+ */
+ public void calculate()
+ {
+ int height = sequences.length;
+
+ total = new Map[maxLength];
+
+ for (int column = start; column <= end; column++)
+ {
+ ResidueCount values = countResidues(column);
+
+ /*
+ * percentage count at or below which we ignore residues
+ */
+ int thresh = (threshold * height) / 100;
+
+ /*
+ * check observed residues in column and record whether each
+ * physico-chemical property is conserved (+1), absence conserved (0),
+ * or not conserved (-1)
+ * Using TreeMap means properties are displayed in alphabetical order
+ */
+ SortedMap<String, Integer> resultHash = new TreeMap<String, Integer>();
+ SymbolCounts symbolCounts = values.getSymbolCounts();
+ char[] symbols = symbolCounts.symbols;
+ int[] counts = symbolCounts.values;
+ for (int j = 0; j < symbols.length; j++)
+ {
+ char c = symbols[j];
+ if (counts[j] > thresh)
+ {
+ recordConservation(resultHash, String.valueOf(c));
+ }
+ }
+ if (values.getGapCount() > thresh)
+ {
+ recordConservation(resultHash, "-");
+ }
+
+ if (total.length > 0)
+ {
+ total[column - start] = resultHash;
+ }
+ }
+ }
+
+ /**
+ * Updates the conservation results for an observed residue
+ *
+ * @param resultMap
+ * a map of {property, conservation} where conservation value is +1
+ * (all residues have the property), 0 (no residue has the property)
+ * or -1 (some do, some don't)
+ * @param res
+ */
+ protected static void recordConservation(Map<String, Integer> resultMap,
+ String res)
+ {
+ res = res.toUpperCase();
+ for (Entry<String, Map<String, Integer>> property : ResidueProperties.propHash
+ .entrySet())
+ {
+ String propertyName = property.getKey();
+ Integer residuePropertyValue = property.getValue().get(res);
+
+ if (!resultMap.containsKey(propertyName))
+ {
+ /*
+ * first time we've seen this residue - note whether it has this property
+ */
+ if (residuePropertyValue != null)
+ {
+ resultMap.put(propertyName, residuePropertyValue);
+ }
+ else
+ {
+ /*
+ * unrecognised residue - use default value for property
+ */
+ resultMap.put(propertyName, property.getValue().get("-"));
+ }
+ }
+ else
+ {
+ Integer currentResult = resultMap.get(propertyName);
+ if (currentResult.intValue() != -1
+ && !currentResult.equals(residuePropertyValue))
+ {
+ /*
+ * property is unconserved - residues seen both with and without it
+ */
+ resultMap.put(propertyName, Integer.valueOf(-1));
+ }
+ }
+ }
+ }
+
+ /**
+ * Counts residues (upper-cased) and gaps in the given column
+ *
+ * @param column
+ * @return
+ */
+ protected ResidueCount countResidues(int column)
+ {
+ ResidueCount values = new ResidueCount(false);
+
+ for (int row = 0; row < sequences.length; row++)
+ {
+ if (sequences[row].getLength() > column)
+ {
+ char c = sequences[row].getCharAt(column);
+ if (canonicaliseAa)
+ {
+ int index = ResidueProperties.aaIndex[c];
+ c = index > 20 ? '-' : ResidueProperties.aa[index].charAt(0);
+ }
+ else
+ {
+ c = toUpperCase(c);
+ }
+ if (Comparison.isGap(c))
+ {
+ values.addGap();
+ }
+ else
+ {
+ values.add(c);
+ }
+ }
+ else
+ {
+ values.addGap();
+ }
+ }
+ return values;
+ }
+
+ /**
+ * Counts conservation and gaps for a column of the alignment
+ *
+ * @return { 1 if fully conserved, else 0, gap count }
+ */
+ public int[] countConservationAndGaps(int column)
+ {
+ int gapCount = 0;
+ boolean fullyConserved = true;
+ int iSize = sequences.length;
+
+ if (iSize == 0)
+ {
+ return new int[] { 0, 0 };
+ }
+
+ char lastRes = '0';
+ for (int i = 0; i < iSize; i++)
+ {
+ if (column >= sequences[i].getLength())
+ {
+ gapCount++;
+ continue;
+ }
+
+ char c = sequences[i].getCharAt(column); // gaps do not have upper/lower case
+
+ if (Comparison.isGap((c)))
+ {
+ gapCount++;
+ }
+ else
+ {
+ c = toUpperCase(c);
+ if (lastRes == '0')
+ {
+ lastRes = c;
+ }
+ if (c != lastRes)
+ {
+ fullyConserved = false;
+ }
+ }
+ }
+
+ int[] r = new int[] { fullyConserved ? 1 : 0, gapCount };
+ return r;
+ }
+
+ /**
+ * Returns the upper-cased character if between 'a' and 'z', else the
+ * unchanged value
+ *
+ * @param c
+ * @return
+ */
+ char toUpperCase(char c)
+ {
+ if ('a' <= c && c <= 'z')
+ {
+ c -= TOUPPERCASE;
+ }
+ return c;
+ }
+
+ /**
+ * Calculates the conservation sequence
+ *
+ * @param positiveOnly
+ * if true, calculate positive conservation; else calculate both
+ * positive and negative conservation
+ * @param maxPercentageGaps
+ * the percentage of gaps in a column, at or above which no
+ * conservation is asserted
+ */
+ public void verdict(boolean positiveOnly, float maxPercentageGaps)
+ {
+ // TODO call this at the end of calculate(), should not be a public method
+
+ StringBuilder consString = new StringBuilder(end);
+
+ // NOTE THIS SHOULD CHECK IF THE CONSEQUENCE ALREADY
+ // EXISTS AND NOT OVERWRITE WITH '-', BUT THIS CASE
+ // DOES NOT EXIST IN JALVIEW 2.1.2
+ for (int i = 0; i < start; i++)
+ {
+ consString.append('-');
+ }
+ consSymbs = new String[end - start + 1];
+ for (int i = start; i <= end; i++)
+ {
+ int[] gapcons = countConservationAndGaps(i);
+ boolean fullyConserved = gapcons[0] == 1;
+ int totGaps = gapcons[1];
+ float pgaps = (totGaps * 100f) / sequences.length;
+
+ if (maxPercentageGaps > pgaps)
+ {
+ Map<String, Integer> resultHash = total[i - start];
+ int count = 0;
+ StringBuilder positives = new StringBuilder(64);
+ StringBuilder negatives = new StringBuilder(32);
+ for (String type : resultHash.keySet())
+ {
+ int result = resultHash.get(type).intValue();
+ if (result == -1)
+ {
+ /*
+ * not conserved (present or absent)
+ */
+ continue;
+ }
+ count++;
+ if (result == 1)
+ {
+ /*
+ * positively conserved property (all residues have it)
+ */
+ positives.append(positives.length() == 0 ? "" : " ");
+ positives.append(type);
+ }
+ if (result == 0 && !positiveOnly)
+ {
+ /*
+ * absense of property is conserved (all residues lack it)
+ */
+ negatives.append(negatives.length() == 0 ? "" : " ");
+ negatives.append("!").append(type);
+ }
+ }
+ if (negatives.length() > 0)
+ {
+ positives.append(" ").append(negatives);
+ }
+ consSymbs[i - start] = positives.toString();
+
+ if (count < 10)
+ {
+ consString.append(count); // Conserved props!=Identity
+ }
+ else
+ {
+ consString.append(fullyConserved ? "*" : "+");
+ }
+ }
+ else
+ {
+ consString.append('-');
+ }
+ }
+
+ consSequence = new Sequence(name, consString.toString(), start, end);
+ }
+
+ /**
+ *
+ *
+ * @return Conservation sequence
+ */
+ public SequenceI getConsSequence()
+ {
+ return consSequence;
+ }
+
+ // From Alignment.java in jalview118
+ public void findQuality()
+ {
+ findQuality(0, maxLength - 1, ScoreModels.getInstance().getBlosum62());
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param sm
+ */
+ private void percentIdentity(ScoreMatrix sm)
+ {
+ seqNums = new Vector<int[]>();
+ int i = 0, iSize = sequences.length;
+ // Do we need to calculate this again?
+ for (i = 0; i < iSize; i++)
+ {
+ calcSeqNum(i, sm);
+ }
+
+ if ((cons2 == null) || seqNumsChanged)
+ {
+ // FIXME remove magic number 24 without changing calc
+ // sm.getSize() returns 25 so doesn't quite do it...
+ cons2 = new int[maxLength][24];
+ cons2GapCounts = new int[maxLength];
+
+ int j = 0;
+
+ while (j < sequences.length)
+ {
+ int[] sqnum = seqNums.elementAt(j);
+
+ for (i = 1; i < sqnum.length; i++)
+ {
+ int index = sqnum[i];
+ if (index == GAP_INDEX)
+ {
+ cons2GapCounts[i - 1]++;
+ }
+ else
+ {
+ cons2[i - 1][index]++;
+ }
+ }
+
+ // TODO should this start from sqnum.length?
+ for (i = sqnum.length - 1; i < maxLength; i++)
+ {
+ cons2GapCounts[i]++;
+ }
+ j++;
+ }
+ }
+ }
+
+ /**
+ * Calculates the quality of the set of sequences over the given inclusive
+ * column range, using the specified substitution score matrix
+ *
+ * @param startCol
+ * @param endCol
+ * @param scoreMatrix
+ */
+ protected void findQuality(int startCol, int endCol, ScoreMatrix scoreMatrix)
+ {
+ quality = new Vector<Double>();
+
+ double max = -Double.MAX_VALUE;
+ float[][] scores = scoreMatrix.getMatrix();
+
+ percentIdentity(scoreMatrix);
+
+ int size = seqNums.size();
+ int[] lengths = new int[size];
+
+ for (int l = 0; l < size; l++)
+ {
+ lengths[l] = seqNums.elementAt(l).length - 1;
+ }
+
+ final int symbolCount = scoreMatrix.getSize();
+
+ for (int j = startCol; j <= endCol; j++)
+ {
+ double bigtot = 0;
+
+ // First Xr = depends on column only
+ double[] x = new double[symbolCount];
+
+ for (int ii = 0; ii < symbolCount; ii++)
+ {
+ x[ii] = 0;
+
+ /*
+ * todo JAL-728 currently assuming last symbol in matrix is * for gap
+ * (which we ignore as counted separately); true for BLOSUM62 but may
+ * not be once alternative matrices are supported
+ */
+ for (int i2 = 0; i2 < symbolCount - 1; i2++)
+ {
+ x[ii] += (((double) cons2[j][i2] * scores[ii][i2]) + 4D);
+ }
+ x[ii] += 4D + cons2GapCounts[j] * scoreMatrix.getMinimumScore();
+
+ x[ii] /= size;
+ }
+
+ // Now calculate D for each position and sum
+ for (int k = 0; k < size; k++)
+ {
+ double tot = 0;
+ double[] xx = new double[symbolCount];
+ // sequence character index, or implied gap if sequence too short
+ int seqNum = (j < lengths[k]) ? seqNums.elementAt(k)[j + 1]
+ : GAP_INDEX;
+
+ for (int i = 0; i < symbolCount - 1; i++)
+ {
+ double sr = 4D;
+ if (seqNum == GAP_INDEX)
+ {
+ sr += scoreMatrix.getMinimumScore();
+ }
+ else
+ {
+ sr += scores[i][seqNum];
+ }
+
+ xx[i] = x[i] - sr;
+
+ tot += (xx[i] * xx[i]);
+ }
+
+ bigtot += Math.sqrt(tot);
+ }
+
+ max = Math.max(max, bigtot);
+
+ quality.addElement(new Double(bigtot));
+ }
+
+ double newmax = -Double.MAX_VALUE;
+
+ for (int j = startCol; j <= endCol; j++)
+ {
+ double tmp = quality.elementAt(j).doubleValue();
+ // tmp = ((max - tmp) * (size - cons2[j][23])) / size;
+ tmp = ((max - tmp) * (size - cons2GapCounts[j])) / size;
+
+ // System.out.println(tmp+ " " + j);
+ quality.setElementAt(new Double(tmp), j);
+
+ if (tmp > newmax)
+ {
+ newmax = tmp;
+ }
+ }
+
+ qualityMinimum = 0D;
+ qualityMaximum = newmax;
+ }
+
+ /**
+ * Complete the given consensus and quuality annotation rows. Note: currently
+ * this method will enlarge the given annotation row if it is too small,
+ * otherwise will leave its length unchanged.
+ *
+ * @param conservation
+ * conservation annotation row
+ * @param quality2
+ * (optional - may be null)
+ * @param istart
+ * first column for conservation
+ * @param alWidth
+ * extent of conservation
+ */
+ public void completeAnnotations(AlignmentAnnotation conservation,
+ AlignmentAnnotation quality2, int istart, int alWidth)
+ {
+ char[] sequence = getConsSequence().getSequence();
+ float minR;
+ float minG;
+ float minB;
+ float maxR;
+ float maxG;
+ float maxB;
+ minR = 0.3f;
+ minG = 0.0f;
+ minB = 0f;
+ maxR = 1.0f - minR;
+ maxG = 0.9f - minG;
+ maxB = 0f - minB; // scalable range for colouring both Conservation and
+ // Quality
+
+ float min = 0f;
+ float max = 11f;
+ float qmin = 0f;
+ float qmax = 0f;
+
+ char c;
+
+ if (conservation != null && conservation.annotations != null
+ && conservation.annotations.length < alWidth)
+ {
+ conservation.annotations = new Annotation[alWidth];
+ }
+
+ if (quality2 != null)
+ {
+ quality2.graphMax = (float) qualityMaximum;
+ if (quality2.annotations != null
+ && quality2.annotations.length < alWidth)
+ {
+ quality2.annotations = new Annotation[alWidth];
+ }
+ qmin = (float) qualityMinimum;
+ qmax = (float) qualityMaximum;
+ }
+
+ for (int i = istart; i < alWidth; i++)
+ {
+ float value = 0;
+
+ c = sequence[i];
+
+ if (Character.isDigit(c))
+ {
+ value = c - '0';
+ }
+ else if (c == '*')
+ {
+ value = 11;
+ }
+ else if (c == '+')
+ {
+ value = 10;
+ }
+
+ if (conservation != null)
+ {
+ float vprop = value - min;
+ vprop /= max;
+ int consp = i - start;
+ String conssym = (value > 0 && consp > -1 && consp < consSymbs.length) ? consSymbs[consp]
+ : "";
+ conservation.annotations[i] = new Annotation(String.valueOf(c),
+ conssym, ' ', value, new Color(minR + (maxR * vprop), minG
+ + (maxG * vprop), minB + (maxB * vprop)));
+ }
+
+ // Quality calc
+ if (quality2 != null)
+ {
+ value = quality.elementAt(i).floatValue();
+ float vprop = value - qmin;
+ vprop /= qmax;
+ quality2.annotations[i] = new Annotation(" ",
+ String.valueOf(value), ' ', value, new Color(minR
+ + (maxR * vprop), minG + (maxG * vprop), minB
+ + (maxB * vprop)));
+ }
+ }
+ }
+
+ /**
+ * construct and call the calculation methods on a new Conservation object
+ *
+ * @param name
+ * - name of conservation
+ * @param seqs
+ * @param start
+ * first column in calculation window
+ * @param end
+ * last column in calculation window
+ * @param positiveOnly
+ * calculate positive (true) or positive and negative (false)
+ * conservation
+ * @param maxPercentGaps
+ * percentage of gaps tolerated in column
+ * @param calcQuality
+ * flag indicating if alignment quality should be calculated
+ * @return Conservation object ready for use in visualization
+ */
+ public static Conservation calculateConservation(String name,
+ List<SequenceI> seqs, int start, int end, boolean positiveOnly,
+ int maxPercentGaps, boolean calcQuality)
+ {
+ Conservation cons = new Conservation(name, seqs, start, end);
+ cons.calculate();
+ cons.verdict(positiveOnly, maxPercentGaps);
+
+ if (calcQuality)
+ {
+ cons.findQuality();
+ }
+
+ return cons;
+ }
+
+ /**
+ * Returns the computed tooltip (annotation description) for a given column.
+ * The tip is empty if the conservation score is zero, otherwise holds the
+ * conserved properties (and, optionally, properties whose absence is
+ * conserved).
+ *
+ * @param column
+ * @return
+ */
+ String getTooltip(int column)
+ {
+ char[] sequence = getConsSequence().getSequence();
+ char val = column < sequence.length ? sequence[column] : '-';
+ boolean hasConservation = val != '-' && val != '0';
+ int consp = column - start;
+ String tip = (hasConservation && consp > -1 && consp < consSymbs.length) ? consSymbs[consp]
+ : "";
+ return tip;
+ }
+}
git://source.jalview.org / jalview.git / blobdiff