- */\r
-package jalview.io;\r
-\r
-import jalview.datamodel.*;\r
-\r
-import jalview.util.*;\r
-\r
-import java.io.*;\r
-\r
-import java.util.*;\r
-\r
-\r
-/**\r
- * DOCUMENT ME!\r
- *\r
- * @author $author$\r
- * @version $Revision$\r
- */\r
-public class JPredFile extends AlignFile\r
-{\r
- Vector ids;\r
- Vector conf;\r
- Hashtable Scores; // Hash of names and score vectors\r
- Hashtable Symscores; // indexes of symbol annotation properties in sequenceI vector\r
- private int QuerySeqPosition;\r
-\r
- /**\r
- * Creates a new JPredFile object.\r
- *\r
- * @param inStr DOCUMENT ME!\r
- */\r
- public JPredFile(String inStr)\r
- {\r
- super(inStr);\r
- }\r
-\r
- /**\r
- * Creates a new JPredFile object.\r
- *\r
- * @param inFile DOCUMENT ME!\r
- * @param type DOCUMENT ME!\r
- *\r
- * @throws IOException DOCUMENT ME!\r
- */\r
- public JPredFile(String inFile, String type) throws IOException\r
- {\r
- super(inFile, type);\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param QuerySeqPosition DOCUMENT ME!\r
- */\r
- public void setQuerySeqPosition(int QuerySeqPosition)\r
- {\r
- this.QuerySeqPosition = QuerySeqPosition;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public int getQuerySeqPosition()\r
- {\r
- return QuerySeqPosition;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public Hashtable getScores()\r
- {\r
- return Scores;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @return DOCUMENT ME!\r
- */\r
- public Hashtable getSymscores()\r
- {\r
- return Symscores;\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- */\r
- public void initData()\r
- {\r
- super.initData();\r
- Scores = new Hashtable();\r
- ids = null;\r
- conf = null;\r
- }\r
-\r
- /**
- * parse a JPred concise file into a sequence-alignment like object.
- */\r
- public void parse() throws IOException\r
- {\r
- // JBPNote log.System.out.println("all read in ");\r
- String line;\r
- QuerySeqPosition = -1;\r
- noSeqs = 0;\r
-\r
- Vector seq_entries = new Vector();\r
- Vector ids = new Vector();\r
- Hashtable Symscores = new Hashtable();\r
-\r
- while ((line = nextLine()) != null)\r
- {\r
- // Concise format allows no comments or non comma-formatted data\r
- StringTokenizer str = new StringTokenizer(line, ":");\r
- String id = "";\r
-\r
- if (!str.hasMoreTokens())\r
- {\r
- continue;\r
- }\r
-\r
- id = str.nextToken();\r
-\r
- String seqsym = str.nextToken();\r
- StringTokenizer symbols = new StringTokenizer(seqsym, ",");\r
-\r
- // decide if we have more than just alphanumeric symbols\r
- int numSymbols = symbols.countTokens();\r
-\r
- if (numSymbols == 0)\r
- {\r
- continue;\r
- }\r
-\r
- if (seqsym.length() != (2 * numSymbols))\r
- {\r
- // Set of scalars for some property\r
- if (Scores.containsKey(id))\r
- {\r
- int i = 1;\r
-\r
- while (Scores.containsKey(id + "_" + i))\r
- {\r
- i++;\r
- }\r
-\r
- id = id + "_" + i;\r
- }\r
-\r
- Vector scores = new Vector();\r
-\r
- // Typecheck from first entry\r
- int i = 0;\r
- String ascore = "dead";\r
-\r
- try\r
- {\r
- // store elements as floats...\r
- while (symbols.hasMoreTokens())\r
- {\r
- ascore = symbols.nextToken();\r
-\r
- Float score = new Float(ascore);\r
- scores.addElement((Object) score);\r
- }\r
-\r
- Scores.put(id, scores);\r
- }\r
- catch (Exception e)\r
- {\r
- // or just keep them as strings\r
- i = scores.size();\r
-\r
- for (int j = 0; j < i; j++)\r
- {\r
- scores.set(j,\r
- (Object) ((Float) scores.get(j)).toString());\r
- }\r
-\r
- scores.addElement((Object) ascore);\r
-\r
- while (symbols.hasMoreTokens())\r
- {\r
- ascore = symbols.nextToken();\r
- scores.addElement((Object) ascore);\r
- }\r
-\r
- Scores.put(id, scores);\r
- }\r
- }\r
- else if (id.equals("jnetconf"))\r
- {\r
- // log.debug System.out.println("here");\r
- id = "Prediction Confidence";\r
- this.conf = new Vector(numSymbols);\r
-\r
- for (int i = 0; i < numSymbols; i++)\r
- {\r
- conf.set(i, (Object) symbols.nextToken());\r
- }\r
- }\r
- else\r
- {\r
- // Sequence or a prediction string (rendered as sequence)\r
- StringBuffer newseq = new StringBuffer();\r
-\r
- for (int i = 0; i < numSymbols; i++)\r
- {\r
- newseq.append(symbols.nextToken());\r
- }\r
-\r
- if (id.indexOf(";") > -1)\r
- {\r
- seq_entries.addElement(newseq);\r
-\r
- int i = 1;\r
- String name = id.substring(id.indexOf(";") + 1);\r
-\r
- while (ids.lastIndexOf(name) > -1)\r
- {\r
- name = id.substring(id.indexOf(";") + 1) + "_" + 1;\r
- }\r
-\r
- ids.addElement(name);\r
-\r
- noSeqs++;\r
- }\r
- else\r
- {\r
- if (id.equals("JNETPRED"))\r
- {\r
- id = "Predicted Secondary Structure";\r
- }\r
-\r
- seq_entries.addElement(newseq.toString());\r
- ids.addElement(id);\r
- Symscores.put((Object) id,\r
- (Object) new Integer(ids.size() - 1));\r
- }\r
- }\r
- }\r
-\r
- if (noSeqs < 1)\r
- {\r
- throw new IOException(\r
- "JpredFile Parser: No sequence in the prediction!");\r
- }\r
-\r
- maxLength = seq_entries.elementAt(0).toString().length();\r
-\r
- for (int i = 0; i < ids.size(); i++)\r
- {\r
- // Add all sequence like objects\r
- Sequence newSeq = new Sequence(ids.elementAt(i).toString(),\r
- seq_entries.elementAt(i).toString(), 1,\r
- seq_entries.elementAt(i).toString().length());\r
-\r
- if (!Symscores.containsKey(ids.elementAt(i)) &&\r
- !isValidProteinSequence(newSeq.getSequence()))\r
- {\r
- throw new IOException(\r
- "JPredConcise: Not a valid protein sequence - (" +\r
- ids.elementAt(i).toString() + ")");\r
- }\r
-\r
- if (maxLength != seq_entries.elementAt(i).toString().length())\r
- {\r
- throw new IOException("JPredConcise: Entry (" +\r
- ids.elementAt(i).toString() +\r
- ") has an unexpected number of columns");\r
- }\r
-\r
- if (newSeq.getName().startsWith("QUERY") &&\r
- (QuerySeqPosition == -1))\r
- {\r
- QuerySeqPosition = seqs.size();\r
- }\r
-\r
- seqs.addElement(newSeq);\r
- }\r
- }\r
-\r
- /**
- * print
- *
- * @return String
- */\r
- public String print()\r
- {\r
- return "Not Supported";\r
- }\r
-\r
- /**\r
- * DOCUMENT ME!\r
- *\r
- * @param args DOCUMENT ME!\r
- */\r
- public static void main(String[] args)\r
- {\r
- try\r
- {\r
- JPredFile blc = new JPredFile(args[0], "File");\r
-\r
- for (int i = 0; i < blc.seqs.size(); i++)\r
- {\r
- System.out.println(((Sequence) blc.seqs.elementAt(i)).getName() +\r
- "\n" + ((Sequence) blc.seqs.elementAt(i)).getSequence() +\r
- "\n");\r
- }\r
- }\r
- catch (java.io.IOException e)\r
- {\r
- System.err.println("Exception " + e);\r
- e.printStackTrace();\r
- }\r
- }\r
-}\r
-\r
-\r
+ */
+package jalview.io;
+
+import java.io.*;
+import java.util.*;
+
+import jalview.datamodel.*;
+
+/**
+ * Parser for the JPred/JNet concise format. This is a series of CSV lines,
+ * each line is either a sequence (QUERY), a sequence profile (align;), or
+ * jnet prediction annotation (anything else).
+ * Automagic translation happens for annotation called 'JNETPRED' (translated to Secondary Structure Prediction), or 'JNETCONF' (translates to 'Prediction Confidence').
+ * Numeric scores are differentiated from symbolic by being parseable into a float vector. They are put in Scores.
+ * Symscores gets the others.
+ * JNetAnnotationMaker translates the data parsed by this object into annotation on an alignment. It is automatically called
+ * but can be used to transfer the annotation onto a sequence in another alignment (and insert gaps where necessary)
+ * @author jprocter
+ * @version $Revision$
+ */
+public class JPredFile
+ extends AlignFile
+{
+ Vector ids;
+ Vector conf;
+ Hashtable Scores; // Hash of names and score vectors
+ Hashtable Symscores; // indexes of symbol annotation properties in sequenceI vector
+ private int QuerySeqPosition;
+
+ /**
+ * Creates a new JPredFile object.
+ *
+ * @param inFile DOCUMENT ME!
+ * @param type DOCUMENT ME!
+ *
+ * @throws IOException DOCUMENT ME!
+ */
+ public JPredFile(String inFile, String type)
+ throws IOException
+ {
+ super(inFile, type);
+ }
+ public JPredFile(FileParse source) throws IOException
+ {
+ super(source);
+ }
+ /**
+ * DOCUMENT ME!
+ *
+ * @param QuerySeqPosition DOCUMENT ME!
+ */
+ public void setQuerySeqPosition(int QuerySeqPosition)
+ {
+ this.QuerySeqPosition = QuerySeqPosition;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public int getQuerySeqPosition()
+ {
+ return QuerySeqPosition;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public Hashtable getScores()
+ {
+ return Scores;
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @return DOCUMENT ME!
+ */
+ public Hashtable getSymscores()
+ {
+ return Symscores;
+ }
+
+ /**
+ * DOCUMENT ME!
+ */
+ public void initData()
+ {
+ super.initData();
+ Scores = new Hashtable();
+ ids = null;
+ conf = null;
+ }
+
+ /**
+ * parse a JPred concise file into a sequence-alignment like object.
+ */
+ public void parse()
+ throws IOException
+ {
+ // JBPNote log.System.out.println("all read in ");
+ String line;
+ QuerySeqPosition = -1;
+ noSeqs = 0;
+
+ Vector seq_entries = new Vector();
+ Vector ids = new Vector();
+ Hashtable Symscores = new Hashtable();
+
+ while ( (line = nextLine()) != null)
+ {
+ // Concise format allows no comments or non comma-formatted data
+ StringTokenizer str = new StringTokenizer(line, ":");
+ String id = "";
+
+ if (!str.hasMoreTokens())
+ {
+ continue;
+ }
+
+ id = str.nextToken();
+
+ String seqsym = str.nextToken();
+ StringTokenizer symbols = new StringTokenizer(seqsym, ",");
+
+ // decide if we have more than just alphanumeric symbols
+ int numSymbols = symbols.countTokens();
+
+ if (numSymbols == 0)
+ {
+ continue;
+ }
+
+ if (seqsym.length() != (2 * numSymbols))
+ {
+ // Set of scalars for some property
+ if (Scores.containsKey(id))
+ {
+ int i = 1;
+
+ while (Scores.containsKey(id + "_" + i))
+ {
+ i++;
+ }
+
+ id = id + "_" + i;
+ }
+
+ Vector scores = new Vector();
+
+ // Typecheck from first entry
+ int i = 0;
+ String ascore = "dead";
+
+ try
+ {
+ // store elements as floats...
+ while (symbols.hasMoreTokens())
+ {
+ ascore = symbols.nextToken();
+
+ Float score = new Float(ascore);
+ scores.addElement( (Object) score);
+ }
+
+ Scores.put(id, scores);
+ }
+ catch (Exception e)
+ {
+ // or just keep them as strings
+ i = scores.size();
+
+ for (int j = 0; j < i; j++)
+ {
+ scores.setElementAt(
+ (Object) ( (Float) scores.elementAt(j)).toString(), j);
+ }
+
+ scores.addElement( (Object) ascore);
+
+ while (symbols.hasMoreTokens())
+ {
+ ascore = symbols.nextToken();
+ scores.addElement( (Object) ascore);
+ }
+
+ Scores.put(id, scores);
+ }
+ }
+ else if (id.equals("jnetconf"))
+ {
+ // log.debug System.out.println("here");
+ id = "Prediction Confidence";
+ this.conf = new Vector(numSymbols);
+
+ for (int i = 0; i < numSymbols; i++)
+ {
+ conf.setElementAt(symbols.nextToken(), i);
+ }
+ }
+ else
+ {
+ // Sequence or a prediction string (rendered as sequence)
+ StringBuffer newseq = new StringBuffer();
+
+ for (int i = 0; i < numSymbols; i++)
+ {
+ newseq.append(symbols.nextToken());
+ }
+
+ if (id.indexOf(";") > -1)
+ {
+ seq_entries.addElement(newseq);
+
+ int i = 1;
+ String name = id.substring(id.indexOf(";") + 1);
+
+ while (ids.lastIndexOf(name) > -1)
+ {
+ name = id.substring(id.indexOf(";") + 1) + "_" + ++i;
+ }
+
+ if (QuerySeqPosition==-1)
+ QuerySeqPosition = ids.size();
+ ids.addElement(name);
+ noSeqs++;
+ }
+ else
+ {
+ if (id.equals("JNETPRED"))
+ {
+ id = "Predicted Secondary Structure";
+ }
+
+ seq_entries.addElement(newseq.toString());
+ ids.addElement(id);
+ Symscores.put( (Object) id,
+ (Object)new Integer(ids.size() - 1));
+ }
+ }
+ }
+ /* leave it to the parser user to actually check this.
+ if (noSeqs < 1)
+ {
+ throw new IOException(
+ "JpredFile Parser: No sequence in the prediction!");
+ }*/
+
+ maxLength = seq_entries.elementAt(0).toString().length();
+
+ for (int i = 0; i < ids.size(); i++)
+ {
+ // Add all sequence like objects
+ Sequence newSeq = new Sequence(ids.elementAt(i).toString(),
+ seq_entries.elementAt(i).toString(), 1,
+ seq_entries.elementAt(i).toString().length());
+
+ if (maxLength != seq_entries.elementAt(i).toString().length())
+ {
+ throw new IOException("JPredConcise: Entry (" +
+ ids.elementAt(i).toString() +
+ ") has an unexpected number of columns");
+ }
+
+ if ((newSeq.getName().startsWith("QUERY") || newSeq.getName().startsWith("align;"))&&
+ (QuerySeqPosition == -1))
+ {
+ QuerySeqPosition = seqs.size();
+ }
+
+ seqs.addElement(newSeq);
+ }
+ if (seqs.size()>0)
+ {
+ // try to make annotation for a prediction only input (default if no alignment is given)
+ Alignment tal = new Alignment(this.getSeqsAsArray());
+ try {
+ JnetAnnotationMaker.add_annotation(this, tal, QuerySeqPosition, true);
+ } catch (Exception e)
+ {
+ tal = null;
+ IOException ex = new IOException("Couldn't parse concise annotation for prediction profile.\n"+e);
+ throw ex;
+ }
+ this.annotations = new Vector();
+ AlignmentAnnotation[] aan = tal.getAlignmentAnnotation();
+ for (int aai = 0; aan!=null && aai<aan.length; aai++)
+ {
+ annotations.addElement(aan[aai]);
+ }
+ }
+ }
+
+ /**
+ * print
+ *
+ * @return String
+ */
+ public String print()
+ {
+ return "Not Supported";
+ }
+
+ /**
+ * DOCUMENT ME!
+ *
+ * @param args DOCUMENT ME!
+ */
+ public static void main(String[] args)
+ {
+ try
+ {
+ JPredFile blc = new JPredFile(args[0], "File");
+
+ for (int i = 0; i < blc.seqs.size(); i++)
+ {
+ System.out.println( ( (Sequence) blc.seqs.elementAt(i)).getName() +
+ "\n" +
+ ( (Sequence) blc.seqs.elementAt(i)).getSequenceAsString() +
+ "\n");
+ }
+ }
+ catch (java.io.IOException e)
+ {
+ System.err.println("Exception " + e);
+ // e.printStackTrace(); not java 1.1 compatible!
+ }
+ }
+
+ Vector annotSeqs = null;
+ /**
+ * removeNonSequences
+ */
+ public void removeNonSequences()
+ {
+ if (annotSeqs != null)
+ {
+ return;
+ }
+ annotSeqs = new Vector();
+ Vector newseqs = new Vector();
+ int i = 0;
+ int j = seqs.size();
+ for (; i < QuerySeqPosition; i++)
+ {
+ annotSeqs.addElement(seqs.elementAt(i));
+ }
+ // check that no stray annotations have been added at the end.
+ {
+ SequenceI sq = (SequenceI) seqs.elementAt(j - 1);
+ if (sq.getName().toUpperCase().startsWith("JPRED"))
+ {
+ annotSeqs.addElement(sq);
+ seqs.removeElementAt(--j);
+ }
+ }
+ for (; i < j; i++)
+ {
+ newseqs.addElement(seqs.elementAt(i));
+ }
+
+ seqs.removeAllElements();
+ seqs = newseqs;
+ }
+}
+