From 0f3f69c44e4a266029ebc16171bebfe34bd5ac96 Mon Sep 17 00:00:00 2001
From: janengelhardt <engelhardt87@googlemail.com>
Date: Wed, 3 Aug 2011 14:30:54 +0200
Subject: [PATCH] JAL-885; Multi-helix bug is solved; One should think about a
 more efficient solution; either by creating a Hashtable of
 pairs in StructureFrequency or directly in a RNA class;

Change-Id: I7bd21eb4d952f22579c786f3d78395949fa046d0
---
 src/jalview/analysis/Rna.java                |   21 +++++++++-
 src/jalview/analysis/StructureFrequency.java |   55 +++++++++++++++-----------
 2 files changed, 52 insertions(+), 24 deletions(-)

diff --git a/src/jalview/analysis/Rna.java b/src/jalview/analysis/Rna.java
index f2ef118..71ccbd1 100644
--- a/src/jalview/analysis/Rna.java
+++ b/src/jalview/analysis/Rna.java
@@ -21,6 +21,7 @@
 
 package jalview.analysis;
 
+import java.util.ArrayList;
 import java.util.Hashtable;
 import java.util.Vector;
 
@@ -28,6 +29,7 @@ import jalview.datamodel.SequenceFeature;
 
 public class Rna
 {
+	static Hashtable<Integer, Integer> pairHash = new Hashtable();
   /**
    * Based off of RALEE code ralee-get-base-pairs. Keeps track of open bracket
    * positions in "stack" vector. When a close bracket is reached, pair this
@@ -62,7 +64,7 @@ public class Rna
         Object temp = stack.lastElement();
         stack.remove(stack.size() - 1);
         pairs.addElement(temp);
-        pairs.addElement(i);
+        pairs.addElement(i);        
       }
 
       i++;
@@ -79,10 +81,27 @@ public class Rna
       
        outPairs[p / 2] = new SequenceFeature("RNA helix", "", "", begin,
               end, "");
+       //pairHash.put(begin, end);
+
     }
 
     return outPairs;
   }
+  
+  
+  /**
+   * Function to get the end position corresponding to a given start position
+   * @param indice - start position of a base pair
+   * @return - end position of a base pair
+   */
+  /*makes no sense at the moment :(
+  public int findEnd(int indice){
+	  //TODO: Probably extend this to find the start to a given end?
+	  //could be done by putting everything twice to the hash
+	  ArrayList<Integer> pair = new ArrayList<Integer>();
+	  return pairHash.get(indice);
+  }*/
+  
 
   /**
    * Figures out which helix each position belongs to and stores the helix
diff --git a/src/jalview/analysis/StructureFrequency.java b/src/jalview/analysis/StructureFrequency.java
index 29c0a0b..a5dd030 100644
--- a/src/jalview/analysis/StructureFrequency.java
+++ b/src/jalview/analysis/StructureFrequency.java
@@ -181,6 +181,15 @@ public class StructureFrequency
     }
   }
   
+  public static int findPair(SequenceFeature[] pairs,int indice){
+	  for(int i=0; i<pairs.length; i++){
+		  if(pairs[i].getBegin()==indice){
+			  return pairs[i].getEnd();
+		  }
+	  }
+	  return -1;
+  }
+  
   /**
    * Method to calculate a 'base pair consensus row', very similar 
    * to nucleotide consensus but takes into account a given structure
@@ -193,7 +202,7 @@ public class StructureFrequency
    */
   public static final void calculate(SequenceI[] sequences, int start,
 		  int end, Hashtable[] result, boolean profile, AlignmentAnnotation rnaStruc){
-
+	  System.out.println("StructureFrequency.calculate");
 	  Hashtable residueHash;
 	  String maxResidue;
 	  char[] seq, struc=rnaStruc.getRNAStruc().toCharArray();
@@ -203,10 +212,15 @@ public class StructureFrequency
 	  int[] values = new int[255];
 	  float percentage;
 	  
+	  /*for(int x=(rna.length-1); x>=0; x--){
+		  System.out.println("BP-"+((rna.length-1)-x)+" "+rna[x].getBegin()+","+rna[x].getEnd()+" "+rna[x].getFeatureGroup());
+	  }*/
+	  
+	  
 	  for (i = start; i < end; i++) //foreach column
 	  {
 		  residueHash = new Hashtable();
-		  maxResidue="";
+		  maxResidue="-";
 		  values = new int[255];
 		  bpEnd=-1;
 
@@ -216,7 +230,7 @@ public class StructureFrequency
 			  s = '-';
 		  }
 
-		  if(s == '-'){
+		  if(s != '('){
 			  values['-']++;
 		  }
 		  else
@@ -245,35 +259,31 @@ public class StructureFrequency
 						  values['-']++;
 						  continue;
 					  }
-					  if(s == '('){
-						  bpEnd=rna[(rna.length-1-nonGap)].getEnd();		  
+					  //if(s == '('){
+						  //bpEnd=rna[(rna.length-1-nonGap)].getEnd();
+						  bpEnd=findPair(rna,i);
 						  cEnd=seq[bpEnd];
+						  //System.out.println(i+","+bpEnd);
 						  if(checkBpType(c,cEnd)){
 							  values['H']++; //H means it's a helix (structured)
 						  }
-					  }
+						  maxResidue="H";
+				  //}
 				  }
 			  }
 			  nonGap++;
 		  }
-		  /*UPDATE this for new values
-	      if (profile)
-	      {
-	        residueHash.put(PROFILE, new int[][]
-	        { values, new int[]
-	        { jSize, nongap } });
-	      }
-		   */
+		  //UPDATE this for new values
+	      //if (profile)
+	      //{
+	     //   residueHash.put(PROFILE, new int[][]
+	      //  { values, new int[]
+	     //   { jSize, nongap } });
+	     // }
+		   
 
 		  count=values['H'];
-		  
-		  if (count == 0)
-		  {
-			  maxResidue = "-";
-		  }else{
-			  maxResidue="H";
-		  }
-		  
+		  		  
 		  residueHash.put(MAXCOUNT, new Integer(count));
 	      residueHash.put(MAXRESIDUE, maxResidue);
 	      
@@ -288,7 +298,6 @@ public class StructureFrequency
 		  if(bpEnd>0){
 			  result[bpEnd]=residueHash;
 		  }
-
 	  }
   }
 
-- 
1.7.10.2