public static final String PROFILE = "P";
- public static final Hashtable[] calculate(Vector sequences, int start,
- int end)
- {
- return calculate(sequences, start, end, false);
- }
+ public static final String PAIRPROFILE = "B";
- public static final Hashtable[] calculate(Vector sequences, int start,
- int end, boolean profile)
+/**
+ * Returns the 3' position of a base pair
+ * @param pairs Secondary structure annotation
+ * @param indice 5' position of a base pair
+ * @return 3' position of a base pair
+ */
+ public static int findPair(SequenceFeature[] pairs, int indice)
{
- SequenceI[] seqs = new SequenceI[sequences.size()];
- int width = 0;
- for (int i = 0; i < sequences.size(); i++)
+ for (int i = 0; i < pairs.length; i++)
{
- seqs[i] = (SequenceI) sequences.elementAt(i);
- if (seqs[i].getLength() > width)
+ if (pairs[i].getBegin() == indice)
{
- width = seqs[i].getLength();
+ return pairs[i].getEnd();
}
}
-
- Hashtable[] reply = new Hashtable[width];
-
- if (end >= width)
- {
- end = width;
- }
-
- calculate(seqs, start, end, reply, profile);
-
- return reply;
- }
-
- public static final void calculate(SequenceI[] sequences, int start,
- int end, Hashtable[] result)
- {
- calculate(sequences, start, end, result, false);
+ return -1;
}
+ /**
+ * Method to calculate a 'base pair consensus row', very similar to nucleotide
+ * consensus but takes into account a given structure
+ *
+ * @param sequences
+ * @param start
+ * @param end
+ * @param result
+ * @param profile
+ * @param rnaStruc
+ */
public static final void calculate(SequenceI[] sequences, int start,
- int end, Hashtable[] result, boolean profile)
+ int end, Hashtable[] result, boolean profile,
+ AlignmentAnnotation rnaStruc)
{
Hashtable residueHash;
- int maxCount, nongap, i, j, v, jSize = sequences.length;
String maxResidue;
- char c;
+ char[] seq, struc = rnaStruc.getRNAStruc().toCharArray();
+ SequenceFeature[] rna = rnaStruc._rnasecstr;
+ char c, s, cEnd;
+ int count, nonGap = 0, i, bpEnd = -1, j, jSize = sequences.length;
+ int[] values;
+ int[][] pairs;
float percentage;
- int[] values = new int[255];
-
- char[] seq;
-
- for (i = start; i < end; i++)
+ for (i = start; i < end; i++) // foreach column
{
residueHash = new Hashtable();
- maxCount = 0;
- maxResidue = "";
- nongap = 0;
+ maxResidue = "-";
values = new int[255];
-
+ pairs = new int[255][255];
+ bpEnd = -1;
+
+ s = struc[i];
+ if (s == '.' || s == ' ')
+ {
+ s = '-';
+ }
- for (j = 0; j < jSize; j++)
+ if (s != '(')
{
- if (sequences[j]==null)
+ if (s == '-')
{
- System.err.println("WARNING: Consensus skipping null sequence - possible race condition.");
- continue;
+ values['-']++;
}
- seq = sequences[j].getSequence();
- if (seq.length > i)
+ }
+ else
+ {
+ for (j = 0; j < jSize; j++) // foreach row
{
- c = seq[i];
-
- if (c == '.' || c == ' ')
- {
- c = '-';
- }
-
- if (c == '-')
+ if (sequences[j] == null)
{
- values['-']++;
+ System.err
+ .println("WARNING: Consensus skipping null sequence - possible race condition.");
continue;
}
- else if ('a' <= c && c <= 'z')
- {
- c -= 32; // ('a' - 'A');
- }
+ seq = sequences[j].getSequence();
- nongap++;
- values[c]++;
+ if (seq.length > i)
+ {
+ c = seq[i];
- }
- else
- {
- values['-']++;
- }
- }
+ // standard representation for gaps in sequence and structure
+ if (c == '.' || c == ' ')
+ {
+ c = '-';
+ }
- for (v = 'A'; v < 'Z'; v++)
- {
- if (values[v] < 2 || values[v] < maxCount)
- {
- continue;
- }
+ if (c == '-')
+ {
+ values['-']++;
+ continue;
+ }
+ bpEnd = findPair(rna, i);
+ cEnd = seq[bpEnd];
+ if (checkBpType(c, cEnd))
+ {
+ values['H']++; // H means it's a helix (structured)
+ }
+ pairs[c][cEnd]++;
- if (values[v] > maxCount)
- {
- maxResidue = String.valueOf((char) v);
- }
- else if (values[v] == maxCount)
- {
- maxResidue += String.valueOf((char) v);
+ maxResidue = "H";
+ }
}
- maxCount = values[v];
- }
-
- if (maxResidue.length() == 0)
- {
- maxResidue = "-";
+ // nonGap++;
}
+ // UPDATE this for new values
if (profile)
{
residueHash.put(PROFILE, new int[][]
{ values, new int[]
- { jSize, nongap } });
+ { jSize, (jSize - values['-']) } });
+
+ residueHash.put(PAIRPROFILE, pairs);
}
- residueHash.put(MAXCOUNT, new Integer(maxCount));
+
+ count = values['H'];
+
+ residueHash.put(MAXCOUNT, new Integer(count));
residueHash.put(MAXRESIDUE, maxResidue);
- percentage = ((float) maxCount * 100) / (float) jSize;
+ percentage = ((float) count * 100) / (float) jSize;
residueHash.put(PID_GAPS, new Float(percentage));
- percentage = ((float) maxCount * 100) / (float) nongap;
- residueHash.put(PID_NOGAPS, new Float(percentage));
- result[i] = residueHash;
+ // percentage = ((float) count * 100) / (float) nongap;
+ // residueHash.put(PID_NOGAPS, new Float(percentage));
+ if (result[i] == null)
+ {
+ result[i] = residueHash;
+ }
+ if (bpEnd > 0)
+ {
+ result[bpEnd] = residueHash;
+ }
}
}
-
- /**
- * Method to calculate a 'base pair consensus row', very similar
- * to nucleotide consensus but takes into account a given structure
- * @param sequences
- * @param start
- * @param end
- * @param result
- * @param profile
- * @param rnaStruc
- */
- public static final void calculate(SequenceI[] sequences, int start,
- int end, Hashtable[] result, boolean profile, AlignmentAnnotation rnaStruc){
-
- Hashtable residueHash;
- String maxResidue;
- char[] seq, struc=rnaStruc.getRNAStruc().toCharArray();
- SequenceFeature[] rna =rnaStruc._rnasecstr;
- char c,s,cEnd;
- int count,nonGap=0,i,bpEnd=-1,j,jSize = sequences.length;
- int[] values = new int[255];
- float percentage;
-
- for (i = start; i < end; i++) //foreach column
- {
- residueHash = new Hashtable();
- maxResidue="";
- values = new int[255];
- bpEnd=-1;
-
- s = struc[i];
- if (s == '.' || s == ' ')
- {
- s = '-';
- }
-
- if(s == '-'){
- values['-']++;
- }
- else
- {
- for (j = 0; j < jSize; j++) //foreach row
- {
- if (sequences[j]==null)
- {
- System.err.println("WARNING: Consensus skipping null sequence - possible race condition.");
- continue;
- }
- seq = sequences[j].getSequence();
-
- if (seq.length > i)
- {
- c = seq[i];
-
- //standard representation for gaps in sequence and structure
- if (c == '.' || c == ' ')
- {
- c = '-';
- }
-
- if (c == '-')
- {
- values['-']++;
- continue;
- }
- if(s == '('){
- bpEnd=rna[(rna.length-1-nonGap)].getEnd();
- cEnd=seq[bpEnd];
- if(checkBpType(c,cEnd)){
- values['H']++; //H means it's a helix (structured)
- }
- }
- }
- }
- 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);
-
- percentage = ((float) count * 100) / (float) jSize;
- residueHash.put(PID_GAPS, new Float(percentage));
-
- //percentage = ((float) count * 100) / (float) nongap;
- //residueHash.put(PID_NOGAPS, new Float(percentage));
- if(result[i]==null){
- result[i] = residueHash;
- }
- if(bpEnd>0){
- result[bpEnd]=residueHash;
- }
-
- }
- }
-
/**
- * Method to check if a base-pair is a canonical or a wobble bp
- * @param up 5' base
- * @param down 3' base
+ * Method to check if a base-pair is a canonical or a wobble bp
+ *
+ * @param up
+ * 5' base
+ * @param down
+ * 3' base
* @return True if it is a canonical/wobble bp
*/
- public static boolean checkBpType(char up, char down){
- if(up>'Z'){up-=32;}
- if(down>'Z'){down-=32;}
-
- switch (up){
- case 'A':
- switch (down){
- case 'T': return true;
- case 'U': return true;
- }
- break;
- case 'C':
- switch (down){
- case 'G': return true;
- }
- break;
- case 'T':
- switch (down){
- case 'A': return true;
- case 'G': return true;
- }
- break;
- case 'G':
- switch (down){
- case 'C': return true;
- case 'T': return true;
- case 'U': return true;
- }
- break;
- case 'U':
- switch (down){
- case 'A': return true;
- case 'G': return true;
- }
- break;
- }
- return false;
+ public static boolean checkBpType(char up, char down)
+ {
+ if (up > 'Z')
+ {
+ up -= 32;
+ }
+ if (down > 'Z')
+ {
+ down -= 32;
+ }
+
+ switch (up)
+ {
+ case 'A':
+ switch (down)
+ {
+ case 'T':
+ return true;
+ case 'U':
+ return true;
+ }
+ break;
+ case 'C':
+ switch (down)
+ {
+ case 'G':
+ return true;
+ }
+ break;
+ case 'T':
+ switch (down)
+ {
+ case 'A':
+ return true;
+ case 'G':
+ return true;
+ }
+ break;
+ case 'G':
+ switch (down)
+ {
+ case 'C':
+ return true;
+ case 'T':
+ return true;
+ case 'U':
+ return true;
+ }
+ break;
+ case 'U':
+ switch (down)
+ {
+ case 'A':
+ return true;
+ case 'G':
+ return true;
+ }
+ break;
+ }
+ return false;
}
-
+
/**
* Compute all or part of the annotation row from the given consensus
* hashtable
{
completeConsensus(consensus, hconsensus, iStart, width,
ignoreGapsInConsensusCalculation, includeAllConsSymbols, null); // new
- // char[]
+ // char[]
// { 'A', 'C', 'G', 'T', 'U' });
}
value = ((Float) hconsensus[i].get(StructureFrequency.PID_GAPS))
.floatValue();
}
-
- String maxRes = hconsensus[i].get(StructureFrequency.MAXRESIDUE).toString();
- String mouseOver = hconsensus[i].get(StructureFrequency.MAXRESIDUE) + " ";
+
+ String maxRes = hconsensus[i].get(StructureFrequency.MAXRESIDUE)
+ .toString();
+ String mouseOver = hconsensus[i].get(StructureFrequency.MAXRESIDUE)
+ + " ";
if (maxRes.length() > 1)
{
mouseOver = "[" + maxRes + "] ";
maxRes = "+";
}
- int[][] profile = (int[][]) hconsensus[i].get(StructureFrequency.PROFILE);
- if (profile != null && includeAllConsSymbols)
+ int[][] profile = (int[][]) hconsensus[i]
+ .get(StructureFrequency.PROFILE);
+ if (profile != null && includeAllConsSymbols) // Just responsible for the
+ // tooltip
{
mouseOver = "";
if (alphabet != null)
}
/**
- * get the sorted profile for the given position of the consensus
+ * get the sorted base-pair profile for the given position of the consensus
*
* @param hconsensus
- * @return
+ * @return profile of the given column
*/
public static int[] extractProfile(Hashtable hconsensus,
- boolean ignoreGapsInConsensusCalculation)
+ boolean ignoreGapsInConsensusCalculation, int column)
{
- int[] rtnval = new int[64];
+ // TODO is there a more elegant way to acces the column number?
+ /*
+ * calculate the frequence of the 16 bp variations for this column 'somehow'
+ * transfer this via getProfile and let it correctly draw
+ */
+ int[] rtnval = new int[51]; // 2*(5*5)+1
int[][] profile = (int[][]) hconsensus.get(StructureFrequency.PROFILE);
+ int[][] pairs = (int[][]) hconsensus
+ .get(StructureFrequency.PAIRPROFILE);
+
if (profile == null)
return null;
- Object[] ca = new Object[profile[0].length];
- float[] vl = new float[profile[0].length];
- for (int c = 0; c < ca.length; c++)
- {
- ca[c] = new char[]
- { (char) c };
- vl[c] = (float) profile[0][c];
- }
- ;
- jalview.util.QuickSort.sort(vl, ca);
+
rtnval[0] = 1;
- for (int c = ca.length - 1; profile[0][((char[]) ca[c])[0]] > 0; c--)
+ for (int j = 65; j <= 90; j++)
{
- if (((char[]) ca[c])[0] != '-')
+ for (int k = 65; k <= 90; k++)
{
- rtnval[rtnval[0]++] = ((char[]) ca[c])[0];
- rtnval[rtnval[0]++] = (int) (((float) profile[0][((char[]) ca[c])[0]]) * 100f / (float) profile[1][ignoreGapsInConsensusCalculation ? 1
- : 0]);
+ if (pairs[j][k] > 0)
+ {
+ rtnval[rtnval[0]++] = j;
+ rtnval[rtnval[0]++] = k;
+ rtnval[rtnval[0]++] = (int) ((float) pairs[j][k] * 100f / (float) profile[1][ignoreGapsInConsensusCalculation ? 1
+ : 0]);
+ }
}
}
+
return rtnval;
}
- enum base {A,T,g,C};
-
-
- public static void main(String args[]){
- //Short test to see if checkBpType works
- ArrayList<String> test = new ArrayList<String>();
- test.add("A");
- test.add("c");
- test.add("g");
- test.add("T");
- test.add("U");
- for (String i : test) {
- for (String j : test) {
- System.out.println(i+"-"+j+": "+StructureFrequency.checkBpType(i.charAt(0),j.charAt(0)));
- }
- }
+ public static void main(String args[])
+ {
+ // Short test to see if checkBpType works
+ ArrayList<String> test = new ArrayList<String>();
+ test.add("A");
+ test.add("c");
+ test.add("g");
+ test.add("T");
+ test.add("U");
+ for (String i : test)
+ {
+ for (String j : test)
+ {
+ System.out.println(i + "-" + j + ": "
+ + StructureFrequency.checkBpType(i.charAt(0), j.charAt(0)));
+ }
+ }
}
}
git://source.jalview.org / jalview.git / blobdiff