-/*\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 jalview.io.NewickFile;\r
-\r
-import jalview.schemes.ResidueProperties;\r
-\r
-import jalview.util.*;\r
-\r
-import java.util.*;\r
-\r
-\r
-/**\r
- * DOCUMENT ME!\r
- *\r
- * @author $author$\r
- * @version $Revision$\r
- */\r
-public class NJTree\r
-{\r
- Vector cluster;\r
- SequenceI[] sequence;\r
-\r
- //SequenceData is a string representation of what the user\r
- //sees. The display may contain hidden columns.\r
- public AlignmentView seqData=null;\r
-\r
- int[] done;\r
- int noseqs;\r
- int noClus;\r
- float[][] distance;\r
- int mini;\r
- int minj;\r
- float ri;\r
- float rj;\r
- Vector groups = new Vector();\r
- SequenceNode maxdist;\r
- SequenceNode top;\r
- float maxDistValue;\r
- float maxheight;\r
- int ycount;\r
- Vector node;\r
- String type;\r
- String pwtype;\r
- Object found = null;\r
- Object leaves = null;\r
-\r
- boolean hasDistances = true; // normal case for jalview trees\r
- boolean hasBootstrap = false; // normal case for jalview trees\r
-\r
- private boolean hasRootDistance = true;\r
-\r
- /**\r
- * Create a new NJTree object with leaves associated with sequences in seqs,\r
- * and original alignment data represented by Cigar strings.\r
- * @param seqs SequenceI[]\r
- * @param odata Cigar[]\r
- * @param treefile NewickFile\r
- */\r
- public NJTree(SequenceI[] seqs, AlignmentView odata, NewickFile treefile) {\r
- this(seqs, treefile);\r
- if (odata!=null)\r
- seqData = odata;\r
- /*\r
- sequenceString = new String[odata.length];\r
- char gapChar = jalview.util.Comparison.GapChars.charAt(0);\r
- for (int i = 0; i < odata.length; i++)\r
- {\r
- SequenceI oseq_aligned = odata[i].getSeq(gapChar);\r
- sequenceString[i] = oseq_aligned.getSequence();\r
- } */\r
- }\r
-\r
- /**\r
- * Creates a new NJTree object from a tree from an external source\r
- *\r
- * @param seqs SequenceI which should be associated with leafs of treefile\r
- * @param treefile A parsed tree\r
- */\r
- public NJTree(SequenceI[] seqs, NewickFile treefile)\r
- {\r
- this.sequence = seqs;\r
- top = treefile.getTree();\r
-\r
- /**\r
- * There is no dependent alignment to be recovered from an\r
- * imported tree.\r
- *\r
- if (sequenceString == null)\r
- {\r
- sequenceString = new String[seqs.length];\r
- for (int i = 0; i < seqs.length; i++)\r
- {\r
- sequenceString[i] = seqs[i].getSequence();\r
- }\r
- }\r
- */\r
-\r
- hasDistances = treefile.HasDistances();\r
- hasBootstrap = treefile.HasBootstrap();\r
- hasRootDistance = treefile.HasRootDistance();\r
-\r
- maxheight = findHeight(top);\r
-\r
- SequenceIdMatcher algnIds = new SequenceIdMatcher(seqs);\r
-\r
- Vector leaves = new Vector();\r
- findLeaves(top, leaves);\r
-\r
- int i = 0;\r
- int namesleft = seqs.length;\r
-\r
- SequenceNode j;\r
- SequenceI nam;\r
- String realnam;\r
-\r
- while (i < leaves.size())\r
- {\r
- j = (SequenceNode) leaves.elementAt(i++);\r
- realnam = j.getName();\r
- nam = null;\r
-\r
- if (namesleft > -1)\r
- {\r
- nam = algnIds.findIdMatch(realnam);\r
- }\r
-\r
- if (nam != null)\r
- {\r
- j.setElement(nam);\r
- namesleft--;\r
- }\r
- else\r
- {\r
- j.setElement(new Sequence(realnam, "THISISAPLACEHLDER"));\r
- j.setPlaceholder(true);\r
- }\r
- }\r
- }\r
-\r
- /**\r
- * Creates a new NJTree object.\r
- *\r
- * @param sequence DOCUMENT ME!\r
- * @param type DOCUMENT ME!\r
- * @param pwtype DOCUMENT ME!\r
- * @param start DOCUMENT ME!\r
- * @param end DOCUMENT ME!\r
- */\r
- public NJTree(SequenceI[] sequence,\r
- AlignmentView seqData,\r
- String type,\r
- String pwtype,\r
- int start, int end)\r
- {\r
- this.sequence = sequence;\r
- this.node = new Vector();\r
- this.type = type;\r
- this.pwtype = pwtype;\r
- if (seqData!=null) {\r
- this.seqData = seqData;\r
- } else {\r
- SeqCigar[] seqs = new SeqCigar[sequence.length];\r
- for(int i=0; i<sequence.length; i++)\r
- {\r
- seqs[i] = new SeqCigar(sequence[i], start, end);\r
- }\r
- CigarArray sdata = new CigarArray(seqs);\r
- sdata.addOperation(CigarArray.M, end-start+1);\r
- this.seqData = new AlignmentView(sdata);\r
- }\r
-\r
- if (!(type.equals("NJ")))\r
- {\r
- type = "AV";\r
- }\r
-\r
- if (!(pwtype.equals("PID")))\r
- {\r
- type = "BL";\r
- }\r
-\r
- int i = 0;\r
-\r
- done = new int[sequence.length];\r
-\r
- while ((i < sequence.length) && (sequence[i] != null))\r
- {\r
- done[i] = 0;\r
- i++;\r
- }\r
-\r
- noseqs = i++;\r
-\r
- distance = findDistances(this.seqData.getSequenceStrings(Comparison.GapChars.charAt(0)));\r
-\r
- makeLeaves();\r
-\r
- noClus = cluster.size();\r
-\r
- cluster();\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public String toString()\r
- {\r
- jalview.io.NewickFile fout = new jalview.io.NewickFile(getTopNode());\r
-\r
- return fout.print(false, true); // distances only\r
- }\r
-\r
- /**\r
- *\r
- * used when the alignment associated to a tree has changed.\r
- *\r
- * @param alignment Vector\r
- */\r
- public void UpdatePlaceHolders(Vector alignment)\r
- {\r
- Vector leaves = new Vector();\r
- findLeaves(top, leaves);\r
-\r
- int sz = leaves.size();\r
- SequenceIdMatcher seqmatcher = null;\r
- int i = 0;\r
-\r
- while (i < sz)\r
- {\r
- SequenceNode leaf = (SequenceNode) leaves.elementAt(i++);\r
-\r
- if (alignment.contains(leaf.element()))\r
- {\r
- leaf.setPlaceholder(false);\r
- }\r
- else\r
- {\r
- if (seqmatcher == null)\r
- {\r
- // Only create this the first time we need it\r
- SequenceI[] seqs = new SequenceI[alignment.size()];\r
-\r
- for (int j = 0; j < seqs.length; j++)\r
- seqs[j] = (SequenceI) alignment.elementAt(j);\r
-\r
- seqmatcher = new SequenceIdMatcher(seqs);\r
- }\r
-\r
- SequenceI nam = seqmatcher.findIdMatch(leaf.getName());\r
-\r
- if (nam != null)\r
- {\r
- leaf.setPlaceholder(false);\r
- leaf.setElement(nam);\r
- }\r
- else\r
- {\r
- leaf.setPlaceholder(true);\r
- }\r
- }\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- */\r
- public void cluster()\r
- {\r
- while (noClus > 2)\r
- {\r
- if (type.equals("NJ"))\r
- {\r
- findMinNJDistance();\r
- }\r
- else\r
- {\r
- findMinDistance();\r
- }\r
-\r
- Cluster c = joinClusters(mini, minj);\r
-\r
- done[minj] = 1;\r
-\r
- cluster.setElementAt(null, minj);\r
- cluster.setElementAt(c, mini);\r
-\r
- noClus--;\r
- }\r
-\r
- boolean onefound = false;\r
-\r
- int one = -1;\r
- int two = -1;\r
-\r
- for (int i = 0; i < noseqs; i++)\r
- {\r
- if (done[i] != 1)\r
- {\r
- if (onefound == false)\r
- {\r
- two = i;\r
- onefound = true;\r
- }\r
- else\r
- {\r
- one = i;\r
- }\r
- }\r
- }\r
-\r
- joinClusters(one, two);\r
- top = (SequenceNode) (node.elementAt(one));\r
-\r
- reCount(top);\r
- findHeight(top);\r
- findMaxDist(top);\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param i DOCUMENT ME!\r
- * @param j DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public Cluster joinClusters(int i, int j)\r
- {\r
- float dist = distance[i][j];\r
-\r
- int noi = ((Cluster) cluster.elementAt(i)).value.length;\r
- int noj = ((Cluster) cluster.elementAt(j)).value.length;\r
-\r
- int[] value = new int[noi + noj];\r
-\r
- for (int ii = 0; ii < noi; ii++)\r
- {\r
- value[ii] = ((Cluster) cluster.elementAt(i)).value[ii];\r
- }\r
-\r
- for (int ii = noi; ii < (noi + noj); ii++)\r
- {\r
- value[ii] = ((Cluster) cluster.elementAt(j)).value[ii - noi];\r
- }\r
-\r
- Cluster c = new Cluster(value);\r
-\r
- ri = findr(i, j);\r
- rj = findr(j, i);\r
-\r
- if (type.equals("NJ"))\r
- {\r
- findClusterNJDistance(i, j);\r
- }\r
- else\r
- {\r
- findClusterDistance(i, j);\r
- }\r
-\r
- SequenceNode sn = new SequenceNode();\r
-\r
- sn.setLeft((SequenceNode) (node.elementAt(i)));\r
- sn.setRight((SequenceNode) (node.elementAt(j)));\r
-\r
- SequenceNode tmpi = (SequenceNode) (node.elementAt(i));\r
- SequenceNode tmpj = (SequenceNode) (node.elementAt(j));\r
-\r
- if (type.equals("NJ"))\r
- {\r
- findNewNJDistances(tmpi, tmpj, dist);\r
- }\r
- else\r
- {\r
- findNewDistances(tmpi, tmpj, dist);\r
- }\r
-\r
- tmpi.setParent(sn);\r
- tmpj.setParent(sn);\r
-\r
- node.setElementAt(sn, i);\r
-\r
- return c;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param tmpi DOCUMENT ME!\r
- * @param tmpj DOCUMENT ME!\r
- * @param dist DOCUMENT ME!\r
- */\r
- public void findNewNJDistances(SequenceNode tmpi, SequenceNode tmpj,\r
- float dist)\r
- {\r
-\r
- tmpi.dist = ((dist + ri) - rj) / 2;\r
- tmpj.dist = (dist - tmpi.dist);\r
-\r
- if (tmpi.dist < 0)\r
- {\r
- tmpi.dist = 0;\r
- }\r
-\r
- if (tmpj.dist < 0)\r
- {\r
- tmpj.dist = 0;\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param tmpi DOCUMENT ME!\r
- * @param tmpj DOCUMENT ME!\r
- * @param dist DOCUMENT ME!\r
- */\r
- public void findNewDistances(SequenceNode tmpi, SequenceNode tmpj,\r
- float dist)\r
- {\r
- float ih = 0;\r
- float jh = 0;\r
-\r
- SequenceNode sni = tmpi;\r
- SequenceNode snj = tmpj;\r
-\r
- while (sni != null)\r
- {\r
- ih = ih + sni.dist;\r
- sni = (SequenceNode) sni.left();\r
- }\r
-\r
- while (snj != null)\r
- {\r
- jh = jh + snj.dist;\r
- snj = (SequenceNode) snj.left();\r
- }\r
-\r
- tmpi.dist = ((dist / 2) - ih);\r
- tmpj.dist = ((dist / 2) - jh);\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param i DOCUMENT ME!\r
- * @param j DOCUMENT ME!\r
- */\r
- public void findClusterDistance(int i, int j)\r
- {\r
- int noi = ((Cluster) cluster.elementAt(i)).value.length;\r
- int noj = ((Cluster) cluster.elementAt(j)).value.length;\r
-\r
- // New distances from cluster to others\r
- float[] newdist = new float[noseqs];\r
-\r
- for (int l = 0; l < noseqs; l++)\r
- {\r
- if ((l != i) && (l != j))\r
- {\r
- newdist[l] = ((distance[i][l] * noi) + (distance[j][l] * noj)) / (noi +\r
- noj);\r
- }\r
- else\r
- {\r
- newdist[l] = 0;\r
- }\r
- }\r
-\r
- for (int ii = 0; ii < noseqs; ii++)\r
- {\r
- distance[i][ii] = newdist[ii];\r
- distance[ii][i] = newdist[ii];\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param i DOCUMENT ME!\r
- * @param j DOCUMENT ME!\r
- */\r
- public void findClusterNJDistance(int i, int j)\r
- {\r
-\r
- // New distances from cluster to others\r
- float[] newdist = new float[noseqs];\r
-\r
- for (int l = 0; l < noseqs; l++)\r
- {\r
- if ((l != i) && (l != j))\r
- {\r
- newdist[l] = ((distance[i][l] + distance[j][l]) -\r
- distance[i][j]) / 2;\r
- }\r
- else\r
- {\r
- newdist[l] = 0;\r
- }\r
- }\r
-\r
- for (int ii = 0; ii < noseqs; ii++)\r
- {\r
- distance[i][ii] = newdist[ii];\r
- distance[ii][i] = newdist[ii];\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param i DOCUMENT ME!\r
- * @param j DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public float findr(int i, int j)\r
- {\r
- float tmp = 1;\r
-\r
- for (int k = 0; k < noseqs; k++)\r
- {\r
- if ((k != i) && (k != j) && (done[k] != 1))\r
- {\r
- tmp = tmp + distance[i][k];\r
- }\r
- }\r
-\r
- if (noClus > 2)\r
- {\r
- tmp = tmp / (noClus - 2);\r
- }\r
-\r
- return tmp;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public float findMinNJDistance()\r
- {\r
- float min = 100000;\r
-\r
- for (int i = 0; i < (noseqs - 1); i++)\r
- {\r
- for (int j = i + 1; j < noseqs; j++)\r
- {\r
- if ((done[i] != 1) && (done[j] != 1))\r
- {\r
- float tmp = distance[i][j] - (findr(i, j) + findr(j, i));\r
-\r
- if (tmp < min)\r
- {\r
- mini = i;\r
- minj = j;\r
-\r
- min = tmp;\r
- }\r
- }\r
- }\r
- }\r
-\r
- return min;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public float findMinDistance()\r
- {\r
- float min = 100000;\r
-\r
- for (int i = 0; i < (noseqs - 1); i++)\r
- {\r
- for (int j = i + 1; j < noseqs; j++)\r
- {\r
- if ((done[i] != 1) && (done[j] != 1))\r
- {\r
- if (distance[i][j] < min)\r
- {\r
- mini = i;\r
- minj = j;\r
-\r
- min = distance[i][j];\r
- }\r
- }\r
- }\r
- }\r
-\r
- return min;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public float[][] findDistances(String[] sequenceString)\r
- {\r
- float[][] distance = new float[noseqs][noseqs];\r
-\r
- if (pwtype.equals("PID"))\r
- {\r
- for (int i = 0; i < (noseqs - 1); i++)\r
- {\r
- for (int j = i; j < noseqs; j++)\r
- {\r
- if (j == i)\r
- {\r
- distance[i][i] = 0;\r
- }\r
- else\r
- {\r
- distance[i][j] = 100 -\r
- Comparison.PID(sequenceString[i], sequenceString[j]);\r
-\r
- distance[j][i] = distance[i][j];\r
- }\r
- }\r
- }\r
- }\r
- else if (pwtype.equals("BL"))\r
- {\r
- int maxscore = 0;\r
- int end = sequenceString[0].length();\r
- for (int i = 0; i < (noseqs - 1); i++)\r
- {\r
- for (int j = i; j < noseqs; j++)\r
- {\r
- int score = 0;\r
-\r
- for (int k = 0; k < end; k++)\r
- {\r
- try\r
- {\r
- score += ResidueProperties.getBLOSUM62(\r
- sequenceString[i].substring(k, k + 1),\r
- sequenceString[j].substring(k, k + 1));\r
- }\r
- catch (Exception ex)\r
- {\r
- System.err.println("err creating BLOSUM62 tree");\r
- ex.printStackTrace();\r
- }\r
- }\r
-\r
- distance[i][j] = (float) score;\r
-\r
- if (score > maxscore)\r
- {\r
- maxscore = score;\r
- }\r
- }\r
- }\r
-\r
- for (int i = 0; i < (noseqs - 1); i++)\r
- {\r
- for (int j = i; j < noseqs; j++)\r
- {\r
- distance[i][j] = (float) maxscore - distance[i][j];\r
- distance[j][i] = distance[i][j];\r
- }\r
- }\r
- }\r
- /* else if (pwtype.equals("SW"))\r
- {\r
- float max = -1;\r
-\r
- for (int i = 0; i < (noseqs - 1); i++)\r
- {\r
- for (int j = i; j < noseqs; j++)\r
- {\r
- AlignSeq as = new AlignSeq(sequence[i], sequence[j], "pep");\r
- as.calcScoreMatrix();\r
- as.traceAlignment();\r
- as.printAlignment(System.out);\r
- distance[i][j] = (float) as.maxscore;\r
-\r
- if (max < distance[i][j])\r
- {\r
- max = distance[i][j];\r
- }\r
- }\r
- }\r
-\r
- for (int i = 0; i < (noseqs - 1); i++)\r
- {\r
- for (int j = i; j < noseqs; j++)\r
- {\r
- distance[i][j] = max - distance[i][j];\r
- distance[j][i] = distance[i][j];\r
- }\r
- }\r
- }/*/\r
-\r
- return distance;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- */\r
- public void makeLeaves()\r
- {\r
- cluster = new Vector();\r
-\r
- for (int i = 0; i < noseqs; i++)\r
- {\r
- SequenceNode sn = new SequenceNode();\r
-\r
- sn.setElement(sequence[i]);\r
- sn.setName(sequence[i].getName());\r
- node.addElement(sn);\r
-\r
- int[] value = new int[1];\r
- value[0] = i;\r
-\r
- Cluster c = new Cluster(value);\r
- cluster.addElement(c);\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- * @param leaves DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public Vector findLeaves(SequenceNode node, Vector leaves)\r
- {\r
- if (node == null)\r
- {\r
- return leaves;\r
- }\r
-\r
- if ((node.left() == null) && (node.right() == null))\r
- {\r
- leaves.addElement(node);\r
-\r
- return leaves;\r
- }\r
- else\r
- {\r
- findLeaves((SequenceNode) node.left(), leaves);\r
- findLeaves((SequenceNode) node.right(), leaves);\r
- }\r
-\r
- return leaves;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- * @param count DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public Object findLeaf(SequenceNode node, int count)\r
- {\r
- found = _findLeaf(node, count);\r
-\r
- return found;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- * @param count DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public Object _findLeaf(SequenceNode node, int count)\r
- {\r
- if (node == null)\r
- {\r
- return null;\r
- }\r
-\r
- if (node.ycount == count)\r
- {\r
- found = node.element();\r
-\r
- return found;\r
- }\r
- else\r
- {\r
- _findLeaf((SequenceNode) node.left(), count);\r
- _findLeaf((SequenceNode) node.right(), count);\r
- }\r
-\r
- return found;\r
- }\r
-\r
- /**\r
- * printNode is mainly for debugging purposes.\r
- *\r
- * @param node SequenceNode\r
- */\r
- public void printNode(SequenceNode node)\r
- {\r
- if (node == null)\r
- {\r
- return;\r
- }\r
-\r
- if ((node.left() == null) && (node.right() == null))\r
- {\r
- System.out.println("Leaf = " +\r
- ((SequenceI) node.element()).getName());\r
- System.out.println("Dist " + ((SequenceNode) node).dist);\r
- System.out.println("Boot " + node.getBootstrap());\r
- }\r
- else\r
- {\r
- System.out.println("Dist " + ((SequenceNode) node).dist);\r
- printNode((SequenceNode) node.left());\r
- printNode((SequenceNode) node.right());\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- */\r
- public void findMaxDist(SequenceNode node)\r
- {\r
- if (node == null)\r
- {\r
- return;\r
- }\r
-\r
- if ((node.left() == null) && (node.right() == null))\r
- {\r
- float dist = ((SequenceNode) node).dist;\r
-\r
- if (dist > maxDistValue)\r
- {\r
- maxdist = (SequenceNode) node;\r
- maxDistValue = dist;\r
- }\r
- }\r
- else\r
- {\r
- findMaxDist((SequenceNode) node.left());\r
- findMaxDist((SequenceNode) node.right());\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public Vector getGroups()\r
- {\r
- return groups;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public float getMaxHeight()\r
- {\r
- return maxheight;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- * @param threshold DOCUMENT ME!\r
- */\r
- public void groupNodes(SequenceNode node, float threshold)\r
- {\r
- if (node == null)\r
- {\r
- return;\r
- }\r
-\r
- if ((node.height / maxheight) > threshold)\r
- {\r
- groups.addElement(node);\r
- }\r
- else\r
- {\r
- groupNodes((SequenceNode) node.left(), threshold);\r
- groupNodes((SequenceNode) node.right(), threshold);\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public float findHeight(SequenceNode node)\r
- {\r
- if (node == null)\r
- {\r
- return maxheight;\r
- }\r
-\r
- if ((node.left() == null) && (node.right() == null))\r
- {\r
- node.height = ((SequenceNode) node.parent()).height + node.dist;\r
-\r
- if (node.height > maxheight)\r
- {\r
- return node.height;\r
- }\r
- else\r
- {\r
- return maxheight;\r
- }\r
- }\r
- else\r
- {\r
- if (node.parent() != null)\r
- {\r
- node.height = ((SequenceNode) node.parent()).height +\r
- node.dist;\r
- }\r
- else\r
- {\r
- maxheight = 0;\r
- node.height = (float) 0.0;\r
- }\r
-\r
- maxheight = findHeight((SequenceNode) (node.left()));\r
- maxheight = findHeight((SequenceNode) (node.right()));\r
- }\r
-\r
- return maxheight;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public SequenceNode reRoot()\r
- {\r
- if (maxdist != null)\r
- {\r
- ycount = 0;\r
-\r
- float tmpdist = maxdist.dist;\r
-\r
- // New top\r
- SequenceNode sn = new SequenceNode();\r
- sn.setParent(null);\r
-\r
- // New right hand of top\r
- SequenceNode snr = (SequenceNode) maxdist.parent();\r
- changeDirection(snr, maxdist);\r
- System.out.println("Printing reversed tree");\r
- printN(snr);\r
- snr.dist = tmpdist / 2;\r
- maxdist.dist = tmpdist / 2;\r
-\r
- snr.setParent(sn);\r
- maxdist.setParent(sn);\r
-\r
- sn.setRight(snr);\r
- sn.setLeft(maxdist);\r
-\r
- top = sn;\r
-\r
- ycount = 0;\r
- reCount(top);\r
- findHeight(top);\r
- }\r
-\r
- return top;\r
- }\r
- /**\r
- *\r
- * @return true if original sequence data can be recovered\r
- */\r
- public boolean hasOriginalSequenceData() {\r
- return seqData!=null;\r
- }\r
- /**\r
- * Returns original alignment data used for calculation - or null where\r
- * not available.\r
- *\r
- * @return null or cut'n'pasteable alignment\r
- */\r
- public String printOriginalSequenceData(char gapChar)\r
- {\r
- if (seqData==null)\r
- return null;\r
-\r
- StringBuffer sb = new StringBuffer();\r
- String[] seqdatas = seqData.getSequenceStrings(gapChar);\r
- for(int i=0; i<seqdatas.length; i++)\r
- {\r
- sb.append(new jalview.util.Format("%-" + 15 + "s").form(\r
- sequence[i].getName()));\r
- sb.append(" "+seqdatas[i]+"\n");\r
- }\r
- return sb.toString();\r
- }\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- */\r
- public void printN(SequenceNode node)\r
- {\r
- if (node == null)\r
- {\r
- return;\r
- }\r
-\r
- if ((node.left() != null) && (node.right() != null))\r
- {\r
- printN((SequenceNode) node.left());\r
- printN((SequenceNode) node.right());\r
- }\r
- else\r
- {\r
- System.out.println(" name = " +\r
- ((SequenceI) node.element()).getName());\r
- }\r
-\r
- System.out.println(" dist = " + ((SequenceNode) node).dist + " " +\r
- ((SequenceNode) node).count + " " + ((SequenceNode) node).height);\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- */\r
- public void reCount(SequenceNode node)\r
- {\r
- ycount = 0;\r
- _reCount(node);\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- */\r
- public void _reCount(SequenceNode node)\r
- {\r
- if (node == null)\r
- {\r
- return;\r
- }\r
-\r
- if ((node.left() != null) && (node.right() != null))\r
- {\r
- _reCount((SequenceNode) node.left());\r
- _reCount((SequenceNode) node.right());\r
-\r
- SequenceNode l = (SequenceNode) node.left();\r
- SequenceNode r = (SequenceNode) node.right();\r
-\r
- ((SequenceNode) node).count = l.count + r.count;\r
- ((SequenceNode) node).ycount = (l.ycount + r.ycount) / 2;\r
- }\r
- else\r
- {\r
- ((SequenceNode) node).count = 1;\r
- ((SequenceNode) node).ycount = ycount++;\r
- }\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- */\r
- public void swapNodes(SequenceNode node)\r
- {\r
- if (node == null)\r
- {\r
- return;\r
- }\r
-\r
- SequenceNode tmp = (SequenceNode) node.left();\r
-\r
- node.setLeft(node.right());\r
- node.setRight(tmp);\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param node DOCUMENT ME!\r
- * @param dir DOCUMENT ME!\r
- */\r
- public void changeDirection(SequenceNode node, SequenceNode dir)\r
- {\r
- if (node == null)\r
- {\r
- return;\r
- }\r
-\r
- if (node.parent() != top)\r
- {\r
- changeDirection((SequenceNode) node.parent(), node);\r
-\r
- SequenceNode tmp = (SequenceNode) node.parent();\r
-\r
- if (dir == node.left())\r
- {\r
- node.setParent(dir);\r
- node.setLeft(tmp);\r
- }\r
- else if (dir == node.right())\r
- {\r
- node.setParent(dir);\r
- node.setRight(tmp);\r
- }\r
- }\r
- else\r
- {\r
- if (dir == node.left())\r
- {\r
- node.setParent(node.left());\r
-\r
- if (top.left() == node)\r
- {\r
- node.setRight(top.right());\r
- }\r
- else\r
- {\r
- node.setRight(top.left());\r
- }\r
- }\r
- else\r
- {\r
- node.setParent(node.right());\r
-\r
- if (top.left() == node)\r
- {\r
- node.setLeft(top.right());\r
- }\r
- else\r
- {\r
- node.setLeft(top.left());\r
- }\r
- }\r
- }\r
- }\r
-\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public SequenceNode getMaxDist()\r
- {\r
- return maxdist;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public SequenceNode getTopNode()\r
- {\r
- return top;\r
- }\r
- /**\r
- *\r
- * @return true if tree has real distances\r
- */\r
- public boolean isHasDistances() {\r
- return hasDistances;\r
- }\r
-\r
- /**\r
- *\r
- * @return true if tree has real bootstrap values\r
- */\r
- public boolean isHasBootstrap() {\r
- return hasBootstrap;\r
- }\r
-\r
- public boolean isHasRootDistance()\r
- {\r
- return hasRootDistance;\r
- }\r
-\r
-}\r
-\r
-\r
-/**\r
- * DOCUMENT ME!\r
- *\r
- * @author $author$\r
- * @version $Revision$\r
- */\r
-class Cluster\r
-{\r
- int[] value;\r
-\r
- /**\r
- * Creates a new Cluster object.\r
- *\r
- * @param value DOCUMENT ME!\r
- */\r
- public Cluster(int[] value)\r
- {\r
- this.value = value;\r
- }\r
-}\r
-\r
+/*
+ * Jalview - A Sequence Alignment Editor and Viewer (Version 2.9)
+ * Copyright (C) 2015 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.api.analysis.ScoreModelI;
+import jalview.datamodel.AlignmentView;
+import jalview.datamodel.BinaryNode;
+import jalview.datamodel.CigarArray;
+import jalview.datamodel.NodeTransformI;
+import jalview.datamodel.SeqCigar;
+import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceI;
+import jalview.datamodel.SequenceNode;
+import jalview.io.NewickFile;
+import jalview.schemes.ResidueProperties;
+
+import java.util.Enumeration;
+import java.util.List;
+import java.util.Vector;
+
+/**
+ * DOCUMENT ME!
+ *
+ * @author $author$
+ * @version $Revision$
+ */
+public class NJTree
+{
+ Vector cluster;
+
+ SequenceI[] sequence;
+
+ // SequenceData is a string representation of what the user
+ // sees. The display may contain hidden columns.
+ public AlignmentView seqData = null;
+
+ int[] done;
+
+ int noseqs;
+
+ int noClus;
+
+ float[][] distance;
+
+ int mini;
+
+ int minj;
+
+ float ri;
+
+ float rj;
+
+ Vector groups = new Vector();
+
+ SequenceNode maxdist;
+
+ SequenceNode top;
+
+ float maxDistValue;
+
+ float maxheight;
+
+ int ycount;
+
+ Vector node;
+
+ String type;
+
+ String pwtype;
+
+ Object found = null;
+
+ Object leaves = null;
+
+ boolean hasDistances = true; // normal case for jalview trees
+
+ boolean hasBootstrap = false; // normal case for jalview trees
+
+ private boolean hasRootDistance = true;
+
+ /**
+ * Create a new NJTree object with leaves associated with sequences in seqs,
+ * and original alignment data represented by Cigar strings.
+ *
+ * @param seqs
+ * SequenceI[]
+ * @param odata
+ * Cigar[]
+ * @param treefile
+ * NewickFile
+ */
+ public NJTree(SequenceI[] seqs, AlignmentView odata, NewickFile treefile)
+ {
+ this(seqs, treefile);
+ if (odata != null)
+ {
+ seqData = odata;
+ }
+ /*
+ * sequenceString = new String[odata.length]; char gapChar =
+ * jalview.util.Comparison.GapChars.charAt(0); for (int i = 0; i <
+ * odata.length; i++) { SequenceI oseq_aligned = odata[i].getSeq(gapChar);
+ * sequenceString[i] = oseq_aligned.getSequence(); }
+ */
+ }
+
+ /**
+ * Creates a new NJTree object from a tree from an external source
+ *
+ * @param seqs
+ * SequenceI which should be associated with leafs of treefile
+ * @param treefile
+ * A parsed tree
+ */
+ public NJTree(SequenceI[] seqs, NewickFile treefile)
+ {
+ this.sequence = seqs;
+ top = treefile.getTree();
+
+ /**
+ * There is no dependent alignment to be recovered from an imported tree.
+ *
+ * if (sequenceString == null) { sequenceString = new String[seqs.length];
+ * for (int i = 0; i < seqs.length; i++) { sequenceString[i] =
+ * seqs[i].getSequence(); } }
+ */
+
+ hasDistances = treefile.HasDistances();
+ hasBootstrap = treefile.HasBootstrap();
+ hasRootDistance = treefile.HasRootDistance();
+
+ maxheight = findHeight(top);
+
+ SequenceIdMatcher algnIds = new SequenceIdMatcher(seqs);
+
+ Vector leaves = new Vector();
+ findLeaves(top, leaves);
+
+ int i = 0;
+ int namesleft = seqs.length;
+
+ SequenceNode j;
+ SequenceI nam;
+ String realnam;
+ Vector one2many = new Vector();
+ int countOne2Many = 0;
+ while (i < leaves.size())
+ {
+ j = (SequenceNode) leaves.elementAt(i++);
+ realnam = j.getName();
+ nam = null;
+
+ if (namesleft > -1)
+ {
+ nam = algnIds.findIdMatch(realnam);
+ }
+
+ if (nam != null)
+ {
+ j.setElement(nam);
+ if (one2many.contains(nam))
+ {
+ countOne2Many++;
+ // if (jalview.bin.Cache.log.isDebugEnabled())
+ // jalview.bin.Cache.log.debug("One 2 many relationship for
+ // "+nam.getName());
+ }
+ else
+ {
+ one2many.addElement(nam);
+ namesleft--;
+ }
+ }
+ else
+ {
+ j.setElement(new Sequence(realnam, "THISISAPLACEHLDER"));
+ j.setPlaceholder(true);
+ }
+ }
+ // if (jalview.bin.Cache.log.isDebugEnabled() && countOne2Many>0) {
+ // jalview.bin.Cache.log.debug("There were "+countOne2Many+" alignment
+ // sequence ids (out of "+one2many.size()+" unique ids) linked to two or
+ // more leaves.");
+ // }
+ // one2many.clear();
+ }
+
+ /**
+ * Creates a new NJTree object.
+ *
+ * @param sequence
+ * DOCUMENT ME!
+ * @param type
+ * DOCUMENT ME!
+ * @param pwtype
+ * DOCUMENT ME!
+ * @param start
+ * DOCUMENT ME!
+ * @param end
+ * DOCUMENT ME!
+ */
+ public NJTree(SequenceI[] sequence, AlignmentView seqData, String type,
+ String pwtype, ScoreModelI sm, int start, int end)
+ {
+ this.sequence = sequence;
+ this.node = new Vector();
+ this.type = type;
+ this.pwtype = pwtype;
+ if (seqData != null)
+ {
+ this.seqData = seqData;
+ }
+ else
+ {
+ SeqCigar[] seqs = new SeqCigar[sequence.length];
+ for (int i = 0; i < sequence.length; i++)
+ {
+ seqs[i] = new SeqCigar(sequence[i], start, end);
+ }
+ CigarArray sdata = new CigarArray(seqs);
+ sdata.addOperation(CigarArray.M, end - start + 1);
+ this.seqData = new AlignmentView(sdata, start);
+ }
+ // System.err.println("Made seqData");// dbg
+ if (!(type.equals("NJ")))
+ {
+ type = "AV";
+ }
+
+ if (sm == null && !(pwtype.equals("PID")))
+ {
+ if (ResidueProperties.getScoreMatrix(pwtype) == null)
+ {
+ pwtype = "BLOSUM62";
+ }
+ }
+
+ int i = 0;
+
+ done = new int[sequence.length];
+
+ while ((i < sequence.length) && (sequence[i] != null))
+ {
+ done[i] = 0;
+ i++;
+ }
+
+ noseqs = i++;
+
+ distance = findDistances(sm);
+ // System.err.println("Made distances");// dbg
+ makeLeaves();
+ // System.err.println("Made leaves");// dbg
+
+ noClus = cluster.size();
+
+ cluster();
+ // System.err.println("Made clusters");// dbg
+
+ }
+
+ /**
+ * Generate a string representation of the Tree
+ *
+ * @return Newick File with all tree data available
+ */
+ public String toString()
+ {
+ jalview.io.NewickFile fout = new jalview.io.NewickFile(getTopNode());
+
+ return fout.print(isHasBootstrap(), isHasDistances(),
+ isHasRootDistance()); // output all data available for tree
+ }
+
+ /**
+ *
+ * used when the alignment associated to a tree has changed.
+ *
+ * @param list
+ * Sequence set to be associated with tree nodes
+ */
+ public void UpdatePlaceHolders(List<SequenceI> list)
+ {
+ Vector leaves = new Vector();
+ findLeaves(top, leaves);
+
+ int sz = leaves.size();
+ SequenceIdMatcher seqmatcher = null;
+ int i = 0;
+
+ while (i < sz)
+ {
+ SequenceNode leaf = (SequenceNode) leaves.elementAt(i++);
+
+ if (list.contains(leaf.element()))
+ {
+ leaf.setPlaceholder(false);
+ }
+ else
+ {
+ if (seqmatcher == null)
+ {
+ // Only create this the first time we need it
+ SequenceI[] seqs = new SequenceI[list.size()];
+
+ for (int j = 0; j < seqs.length; j++)
+ {
+ seqs[j] = list.get(j);
+ }
+
+ seqmatcher = new SequenceIdMatcher(seqs);
+ }
+
+ SequenceI nam = seqmatcher.findIdMatch(leaf.getName());
+
+ if (nam != null)
+ {
+ if (!leaf.isPlaceholder())
+ {
+ // remapping the node to a new sequenceI - should remove any refs to
+ // old one.
+ // TODO - make many sequenceI to one leaf mappings possible!
+ // (JBPNote)
+ }
+ leaf.setPlaceholder(false);
+ leaf.setElement(nam);
+ }
+ else
+ {
+ if (!leaf.isPlaceholder())
+ {
+ // Construct a new placeholder sequence object for this leaf
+ leaf.setElement(new Sequence(leaf.getName(),
+ "THISISAPLACEHLDER"));
+ }
+ leaf.setPlaceholder(true);
+
+ }
+ }
+ }
+ }
+
+ /**
+ * rename any nodes according to their associated sequence. This will modify
+ * the tree's metadata! (ie the original NewickFile or newly generated
+ * BinaryTree's label data)
+ */
+ public void renameAssociatedNodes()
+ {
+ applyToNodes(new NodeTransformI()
+ {
+
+ @Override
+ public void transform(BinaryNode node)
+ {
+ Object el = node.element();
+ if (el != null && el instanceof SequenceI)
+ {
+ node.setName(((SequenceI) el).getName());
+ }
+ }
+ });
+ }
+
+ /**
+ * DOCUMENT ME!
+ */
+ public void cluster()
+ {
+ while (noClus > 2)
+ {
+ if (type.equals("NJ"))
+ {
+ findMinNJDistance();
+ }
+ else
+ {
+ findMinDistance();
+ }
+
+ Cluster c = joinClusters(mini, minj);
+
+ done[minj] = 1;
+
+ cluster.setElementAt(null, minj);
+ cluster.setElementAt(c, mini);
+
+ noClus--;
+ }
+
+ boolean onefound = false;
+
+ int one = -1;
+ int two = -1;
+
+ for (int i = 0; i < noseqs; i++)
+ {
+ if (done[i] != 1)
+ {
+ if (onefound == false)
+ {
+ two = i;
+ onefound = true;
+ }
+ else
+ {
+ one = i;
+ }
+ }
+ }
+
+ joinClusters(one, two);
+ top = (SequenceNode) (node.elementAt(one));
+
+ reCount(top);
+ findHeight(top);
+ findMaxDist(top);
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param i
+ * DOCUMENT ME!
+ * @param j
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public Cluster joinClusters(int i, int j)
+ {
+ float dist = distance[i][j];
+
+ int noi = ((Cluster) cluster.elementAt(i)).value.length;
+ int noj = ((Cluster) cluster.elementAt(j)).value.length;
+
+ int[] value = new int[noi + noj];
+
+ for (int ii = 0; ii < noi; ii++)
+ {
+ value[ii] = ((Cluster) cluster.elementAt(i)).value[ii];
+ }
+
+ for (int ii = noi; ii < (noi + noj); ii++)
+ {
+ value[ii] = ((Cluster) cluster.elementAt(j)).value[ii - noi];
+ }
+
+ Cluster c = new Cluster(value);
+
+ ri = findr(i, j);
+ rj = findr(j, i);
+
+ if (type.equals("NJ"))
+ {
+ findClusterNJDistance(i, j);
+ }
+ else
+ {
+ findClusterDistance(i, j);
+ }
+
+ SequenceNode sn = new SequenceNode();
+
+ sn.setLeft((SequenceNode) (node.elementAt(i)));
+ sn.setRight((SequenceNode) (node.elementAt(j)));
+
+ SequenceNode tmpi = (SequenceNode) (node.elementAt(i));
+ SequenceNode tmpj = (SequenceNode) (node.elementAt(j));
+
+ if (type.equals("NJ"))
+ {
+ findNewNJDistances(tmpi, tmpj, dist);
+ }
+ else
+ {
+ findNewDistances(tmpi, tmpj, dist);
+ }
+
+ tmpi.setParent(sn);
+ tmpj.setParent(sn);
+
+ node.setElementAt(sn, i);
+
+ return c;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param tmpi
+ * DOCUMENT ME!
+ * @param tmpj
+ * DOCUMENT ME!
+ * @param dist
+ * DOCUMENT ME!
+ */
+ public void findNewNJDistances(SequenceNode tmpi, SequenceNode tmpj,
+ float dist)
+ {
+
+ tmpi.dist = ((dist + ri) - rj) / 2;
+ tmpj.dist = (dist - tmpi.dist);
+
+ if (tmpi.dist < 0)
+ {
+ tmpi.dist = 0;
+ }
+
+ if (tmpj.dist < 0)
+ {
+ tmpj.dist = 0;
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param tmpi
+ * DOCUMENT ME!
+ * @param tmpj
+ * DOCUMENT ME!
+ * @param dist
+ * DOCUMENT ME!
+ */
+ public void findNewDistances(SequenceNode tmpi, SequenceNode tmpj,
+ float dist)
+ {
+ float ih = 0;
+ float jh = 0;
+
+ SequenceNode sni = tmpi;
+ SequenceNode snj = tmpj;
+
+ while (sni != null)
+ {
+ ih = ih + sni.dist;
+ sni = (SequenceNode) sni.left();
+ }
+
+ while (snj != null)
+ {
+ jh = jh + snj.dist;
+ snj = (SequenceNode) snj.left();
+ }
+
+ tmpi.dist = ((dist / 2) - ih);
+ tmpj.dist = ((dist / 2) - jh);
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param i
+ * DOCUMENT ME!
+ * @param j
+ * DOCUMENT ME!
+ */
+ public void findClusterDistance(int i, int j)
+ {
+ int noi = ((Cluster) cluster.elementAt(i)).value.length;
+ int noj = ((Cluster) cluster.elementAt(j)).value.length;
+
+ // New distances from cluster to others
+ float[] newdist = new float[noseqs];
+
+ for (int l = 0; l < noseqs; l++)
+ {
+ if ((l != i) && (l != j))
+ {
+ newdist[l] = ((distance[i][l] * noi) + (distance[j][l] * noj))
+ / (noi + noj);
+ }
+ else
+ {
+ newdist[l] = 0;
+ }
+ }
+
+ for (int ii = 0; ii < noseqs; ii++)
+ {
+ distance[i][ii] = newdist[ii];
+ distance[ii][i] = newdist[ii];
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param i
+ * DOCUMENT ME!
+ * @param j
+ * DOCUMENT ME!
+ */
+ public void findClusterNJDistance(int i, int j)
+ {
+
+ // New distances from cluster to others
+ float[] newdist = new float[noseqs];
+
+ for (int l = 0; l < noseqs; l++)
+ {
+ if ((l != i) && (l != j))
+ {
+ newdist[l] = ((distance[i][l] + distance[j][l]) - distance[i][j]) / 2;
+ }
+ else
+ {
+ newdist[l] = 0;
+ }
+ }
+
+ for (int ii = 0; ii < noseqs; ii++)
+ {
+ distance[i][ii] = newdist[ii];
+ distance[ii][i] = newdist[ii];
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param i
+ * DOCUMENT ME!
+ * @param j
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public float findr(int i, int j)
+ {
+ float tmp = 1;
+
+ for (int k = 0; k < noseqs; k++)
+ {
+ if ((k != i) && (k != j) && (done[k] != 1))
+ {
+ tmp = tmp + distance[i][k];
+ }
+ }
+
+ if (noClus > 2)
+ {
+ tmp = tmp / (noClus - 2);
+ }
+
+ return tmp;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public float findMinNJDistance()
+ {
+ float min = 100000;
+
+ for (int i = 0; i < (noseqs - 1); i++)
+ {
+ for (int j = i + 1; j < noseqs; j++)
+ {
+ if ((done[i] != 1) && (done[j] != 1))
+ {
+ float tmp = distance[i][j] - (findr(i, j) + findr(j, i));
+
+ if (tmp < min)
+ {
+ mini = i;
+ minj = j;
+
+ min = tmp;
+ }
+ }
+ }
+ }
+
+ return min;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public float findMinDistance()
+ {
+ float min = 100000;
+
+ for (int i = 0; i < (noseqs - 1); i++)
+ {
+ for (int j = i + 1; j < noseqs; j++)
+ {
+ if ((done[i] != 1) && (done[j] != 1))
+ {
+ if (distance[i][j] < min)
+ {
+ mini = i;
+ minj = j;
+
+ min = distance[i][j];
+ }
+ }
+ }
+ }
+
+ return min;
+ }
+
+ /**
+ * Calculate a distance matrix given the sequence input data and score model
+ *
+ * @return similarity matrix used to compute tree
+ */
+ public float[][] findDistances(ScoreModelI _pwmatrix)
+ {
+
+ float[][] distance = new float[noseqs][noseqs];
+ if (_pwmatrix == null)
+ {
+ // Resolve substitution model
+ _pwmatrix = ResidueProperties.getScoreModel(pwtype);
+ if (_pwmatrix == null)
+ {
+ _pwmatrix = ResidueProperties.getScoreMatrix("BLOSUM62");
+ }
+ }
+ distance = _pwmatrix.findDistances(seqData);
+ return distance;
+
+ }
+
+ /**
+ * DOCUMENT ME!
+ */
+ public void makeLeaves()
+ {
+ cluster = new Vector();
+
+ for (int i = 0; i < noseqs; i++)
+ {
+ SequenceNode sn = new SequenceNode();
+
+ sn.setElement(sequence[i]);
+ sn.setName(sequence[i].getName());
+ node.addElement(sn);
+
+ int[] value = new int[1];
+ value[0] = i;
+
+ Cluster c = new Cluster(value);
+ cluster.addElement(c);
+ }
+ }
+
+ /**
+ * Search for leaf nodes.
+ *
+ * @param node
+ * root node to search from
+ * @param leaves
+ * Vector of leaves to add leaf node objects too.
+ *
+ * @return Vector of leaf nodes on binary tree
+ */
+ public Vector findLeaves(SequenceNode node, Vector leaves)
+ {
+ if (node == null)
+ {
+ return leaves;
+ }
+
+ if ((node.left() == null) && (node.right() == null)) // Interior node
+ // detection
+ {
+ leaves.addElement(node);
+
+ return leaves;
+ }
+ else
+ {
+ /*
+ * TODO: Identify internal nodes... if (node.isSequenceLabel()) {
+ * leaves.addElement(node); }
+ */
+ findLeaves((SequenceNode) node.left(), leaves);
+ findLeaves((SequenceNode) node.right(), leaves);
+ }
+
+ return leaves;
+ }
+
+ /**
+ * Find the leaf node with a particular ycount
+ *
+ * @param node
+ * initial point on tree to search from
+ * @param count
+ * value to search for
+ *
+ * @return null or the node with ycound=count
+ */
+ public Object findLeaf(SequenceNode node, int count)
+ {
+ found = _findLeaf(node, count);
+
+ return found;
+ }
+
+ /*
+ * #see findLeaf(SequenceNode node, count)
+ */
+ public Object _findLeaf(SequenceNode node, int count)
+ {
+ if (node == null)
+ {
+ return null;
+ }
+
+ if (node.ycount == count)
+ {
+ found = node.element();
+
+ return found;
+ }
+ else
+ {
+ _findLeaf((SequenceNode) node.left(), count);
+ _findLeaf((SequenceNode) node.right(), count);
+ }
+
+ return found;
+ }
+
+ /**
+ * printNode is mainly for debugging purposes.
+ *
+ * @param node
+ * SequenceNode
+ */
+ public void printNode(SequenceNode node)
+ {
+ if (node == null)
+ {
+ return;
+ }
+
+ if ((node.left() == null) && (node.right() == null))
+ {
+ System.out
+ .println("Leaf = " + ((SequenceI) node.element()).getName());
+ System.out.println("Dist " + node.dist);
+ System.out.println("Boot " + node.getBootstrap());
+ }
+ else
+ {
+ System.out.println("Dist " + node.dist);
+ printNode((SequenceNode) node.left());
+ printNode((SequenceNode) node.right());
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ */
+ public void findMaxDist(SequenceNode node)
+ {
+ if (node == null)
+ {
+ return;
+ }
+
+ if ((node.left() == null) && (node.right() == null))
+ {
+ float dist = node.dist;
+
+ if (dist > maxDistValue)
+ {
+ maxdist = node;
+ maxDistValue = dist;
+ }
+ }
+ else
+ {
+ findMaxDist((SequenceNode) node.left());
+ findMaxDist((SequenceNode) node.right());
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public Vector getGroups()
+ {
+ return groups;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public float getMaxHeight()
+ {
+ return maxheight;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ * @param threshold
+ * DOCUMENT ME!
+ */
+ public void groupNodes(SequenceNode node, float threshold)
+ {
+ if (node == null)
+ {
+ return;
+ }
+
+ if ((node.height / maxheight) > threshold)
+ {
+ groups.addElement(node);
+ }
+ else
+ {
+ groupNodes((SequenceNode) node.left(), threshold);
+ groupNodes((SequenceNode) node.right(), threshold);
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public float findHeight(SequenceNode node)
+ {
+ if (node == null)
+ {
+ return maxheight;
+ }
+
+ if ((node.left() == null) && (node.right() == null))
+ {
+ node.height = ((SequenceNode) node.parent()).height + node.dist;
+
+ if (node.height > maxheight)
+ {
+ return node.height;
+ }
+ else
+ {
+ return maxheight;
+ }
+ }
+ else
+ {
+ if (node.parent() != null)
+ {
+ node.height = ((SequenceNode) node.parent()).height + node.dist;
+ }
+ else
+ {
+ maxheight = 0;
+ node.height = (float) 0.0;
+ }
+
+ maxheight = findHeight((SequenceNode) (node.left()));
+ maxheight = findHeight((SequenceNode) (node.right()));
+ }
+
+ return maxheight;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public SequenceNode reRoot()
+ {
+ if (maxdist != null)
+ {
+ ycount = 0;
+
+ float tmpdist = maxdist.dist;
+
+ // New top
+ SequenceNode sn = new SequenceNode();
+ sn.setParent(null);
+
+ // New right hand of top
+ SequenceNode snr = (SequenceNode) maxdist.parent();
+ changeDirection(snr, maxdist);
+ System.out.println("Printing reversed tree");
+ printN(snr);
+ snr.dist = tmpdist / 2;
+ maxdist.dist = tmpdist / 2;
+
+ snr.setParent(sn);
+ maxdist.setParent(sn);
+
+ sn.setRight(snr);
+ sn.setLeft(maxdist);
+
+ top = sn;
+
+ ycount = 0;
+ reCount(top);
+ findHeight(top);
+ }
+
+ return top;
+ }
+
+ /**
+ *
+ * @return true if original sequence data can be recovered
+ */
+ public boolean hasOriginalSequenceData()
+ {
+ return seqData != null;
+ }
+
+ /**
+ * Returns original alignment data used for calculation - or null where not
+ * available.
+ *
+ * @return null or cut'n'pasteable alignment
+ */
+ public String printOriginalSequenceData(char gapChar)
+ {
+ if (seqData == null)
+ {
+ return null;
+ }
+
+ StringBuffer sb = new StringBuffer();
+ String[] seqdatas = seqData.getSequenceStrings(gapChar);
+ for (int i = 0; i < seqdatas.length; i++)
+ {
+ sb.append(new jalview.util.Format("%-" + 15 + "s").form(sequence[i]
+ .getName()));
+ sb.append(" " + seqdatas[i] + "\n");
+ }
+ return sb.toString();
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ */
+ public void printN(SequenceNode node)
+ {
+ if (node == null)
+ {
+ return;
+ }
+
+ if ((node.left() != null) && (node.right() != null))
+ {
+ printN((SequenceNode) node.left());
+ printN((SequenceNode) node.right());
+ }
+ else
+ {
+ System.out.println(" name = "
+ + ((SequenceI) node.element()).getName());
+ }
+
+ System.out.println(" dist = " + node.dist + " " + node.count + " "
+ + node.height);
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ */
+ public void reCount(SequenceNode node)
+ {
+ ycount = 0;
+ _lycount = 0;
+ // _lylimit = this.node.size();
+ _reCount(node);
+ }
+
+ private long _lycount = 0, _lylimit = 0;
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ */
+ public void _reCount(SequenceNode node)
+ {
+ // if (_lycount<_lylimit)
+ // {
+ // System.err.println("Warning: depth of _recount greater than number of nodes.");
+ // }
+ if (node == null)
+ {
+ return;
+ }
+ _lycount++;
+
+ if ((node.left() != null) && (node.right() != null))
+ {
+
+ _reCount((SequenceNode) node.left());
+ _reCount((SequenceNode) node.right());
+
+ SequenceNode l = (SequenceNode) node.left();
+ SequenceNode r = (SequenceNode) node.right();
+
+ node.count = l.count + r.count;
+ node.ycount = (l.ycount + r.ycount) / 2;
+ }
+ else
+ {
+ node.count = 1;
+ node.ycount = ycount++;
+ }
+ _lycount--;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ */
+ public void swapNodes(SequenceNode node)
+ {
+ if (node == null)
+ {
+ return;
+ }
+
+ SequenceNode tmp = (SequenceNode) node.left();
+
+ node.setLeft(node.right());
+ node.setRight(tmp);
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param node
+ * DOCUMENT ME!
+ * @param dir
+ * DOCUMENT ME!
+ */
+ public void changeDirection(SequenceNode node, SequenceNode dir)
+ {
+ if (node == null)
+ {
+ return;
+ }
+
+ if (node.parent() != top)
+ {
+ changeDirection((SequenceNode) node.parent(), node);
+
+ SequenceNode tmp = (SequenceNode) node.parent();
+
+ if (dir == node.left())
+ {
+ node.setParent(dir);
+ node.setLeft(tmp);
+ }
+ else if (dir == node.right())
+ {
+ node.setParent(dir);
+ node.setRight(tmp);
+ }
+ }
+ else
+ {
+ if (dir == node.left())
+ {
+ node.setParent(node.left());
+
+ if (top.left() == node)
+ {
+ node.setRight(top.right());
+ }
+ else
+ {
+ node.setRight(top.left());
+ }
+ }
+ else
+ {
+ node.setParent(node.right());
+
+ if (top.left() == node)
+ {
+ node.setLeft(top.right());
+ }
+ else
+ {
+ node.setLeft(top.left());
+ }
+ }
+ }
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public SequenceNode getMaxDist()
+ {
+ return maxdist;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public SequenceNode getTopNode()
+ {
+ return top;
+ }
+
+ /**
+ *
+ * @return true if tree has real distances
+ */
+ public boolean isHasDistances()
+ {
+ return hasDistances;
+ }
+
+ /**
+ *
+ * @return true if tree has real bootstrap values
+ */
+ public boolean isHasBootstrap()
+ {
+ return hasBootstrap;
+ }
+
+ public boolean isHasRootDistance()
+ {
+ return hasRootDistance;
+ }
+
+ /**
+ * apply the given transform to all the nodes in the tree.
+ *
+ * @param nodeTransformI
+ */
+ public void applyToNodes(NodeTransformI nodeTransformI)
+ {
+ for (Enumeration nodes = node.elements(); nodes.hasMoreElements(); nodeTransformI
+ .transform((BinaryNode) nodes.nextElement()))
+ {
+ ;
+ }
+ }
+}
+
+/**
+ * DOCUMENT ME!
+ *
+ * @author $author$
+ * @version $Revision$
+ */
+class Cluster
+{
+ int[] value;
+
+ /**
+ * Creates a new Cluster object.
+ *
+ * @param value
+ * DOCUMENT ME!
+ */
+ public Cluster(int[] value)
+ {
+ this.value = value;
+ }
+}
git://source.jalview.org / jalview.git / blobdiff