* The alignment column on which the first profile is based.
* @param end
* The alignment column on which the last profile is based.
- * @param saveFullProfile
- * if true, all residue counts are saved (enables profile logo)
* @param removeBelowBackground
* if true, symbols with a match emission probability less than
* background frequency are ignored
* @return
*/
public static ProfilesI calculateHMMProfiles(final HiddenMarkovModel hmm,
- int width, int start, int end, boolean saveFullProfile,
- boolean removeBelowBackground, boolean infoLetterHeight)
+ int width, int start, int end, boolean removeBelowBackground,
+ boolean infoLetterHeight)
{
ProfileI[] result = new ProfileI[width];
char[] symbols = hmm.getSymbols().toCharArray();
int gapCount = counts.getGapCount();
ProfileI profile = new Profile(symbolCount, gapCount, maxCount,
maxResidue);
-
- if (saveFullProfile)
- {
- profile.setCounts(counts);
- }
+ profile.setCounts(counts);
result[column] = profile;
}
* @param endCol
* end column (exclusive)
* @param ignoreGaps
- * if true, normalise residue percentages
+ * if true, normalise residue percentages
* @param showSequenceLogo
* if true include all information symbols, else just show modal
* residue
- * @param nseq
- * number of sequences
*/
public static float completeInformation(AlignmentAnnotation information,
- ProfilesI profiles, int startCol, int endCol, long nseq,
- float currentMax)
+ ProfilesI profiles, int startCol, int endCol)
{
// long now = System.currentTimeMillis();
- if (information == null || information.annotations == null
- || information.annotations.length < endCol)
+ if (information == null || information.annotations == null)
{
/*
* called with a bad alignment annotation row
float max = 0f;
SequenceI hmmSeq = information.sequenceRef;
+
+ int seqLength = hmmSeq.getLength();
+ if (information.annotations.length < seqLength)
+ {
+ return 0;
+ }
+
HiddenMarkovModel hmm = hmmSeq.getHMM();
for (int column = startCol; column < endCol; column++)
{
- ProfileI profile = profiles.get(column);
- if (profile == null)
+ if (column >= seqLength)
{
- /*
- * happens if sequences calculated over were
- * shorter than alignment width
- */
- information.annotations[column] = null;
- return 0f;
+ // hmm consensus sequence is shorter than the alignment
+ break;
}
float value = hmm.getInformationContent(column);
String description = isNaN ? null
: String.format("%.4f bits", value);
information.annotations[column] = new Annotation(
- Character.toString(Character
- .toUpperCase(hmm.getConsensusAtAlignColumn(column))),
+ Character.toString(
+ Character.toUpperCase(hmmSeq.getCharAt(column))),
description, ' ', value);
}
- max = Math.max(max, currentMax);
information.graphMax = max;
return max;
}
* @param endCol
* end column (exclusive)
*/
- public static void completeOccupancyAnnot(AlignmentAnnotation occupancy,
+ public static void completeGapAnnot(AlignmentAnnotation occupancy,
ProfilesI profiles, int startCol, int endCol, long nseq)
{
if (occupancy == null || occupancy.annotations == null
* contains
*
* <pre>
- * [profileType, numberOfValues, nonGapCount, charValue1, percentage1, charValue2, percentage2, ...]
+ * [profileType, numberOfValues, totalPercent, charValue1, percentage1, charValue2, percentage2, ...]
* in descending order of percentage value
* </pre>
*
*/
public static int[] extractProfile(ProfileI profile, boolean ignoreGaps)
{
- int[] rtnval = new int[64];
ResidueCount counts = profile.getCounts();
if (counts == null)
{
char[] symbols = symbolCounts.symbols;
int[] values = symbolCounts.values;
QuickSort.sort(values, symbols);
- int nextArrayPos = 2;
int totalPercentage = 0;
final int divisor = ignoreGaps ? profile.getNonGapped()
: profile.getHeight();
/*
* traverse the arrays in reverse order (highest counts first)
*/
+ int[] result = new int[3 + 2 * symbols.length];
+ int nextArrayPos = 3;
+ int nonZeroCount = 0;
+
for (int i = symbols.length - 1; i >= 0; i--)
{
int theChar = symbols[i];
int charCount = values[i];
-
- rtnval[nextArrayPos++] = theChar;
final int percentage = (charCount * 100) / divisor;
- rtnval[nextArrayPos++] = percentage;
+ if (percentage == 0)
+ {
+ /*
+ * this count (and any remaining) round down to 0% - discard
+ */
+ break;
+ }
+ nonZeroCount++;
+ result[nextArrayPos++] = theChar;
+ result[nextArrayPos++] = percentage;
totalPercentage += percentage;
}
- rtnval[0] = symbols.length;
- rtnval[1] = totalPercentage;
- int[] result = new int[rtnval.length + 1];
+
+ /*
+ * truncate array if any zero values were discarded
+ */
+ if (nonZeroCount < symbols.length)
+ {
+ int[] tmp = new int[3 + 2 * nonZeroCount];
+ System.arraycopy(result, 0, tmp, 0, tmp.length);
+ result = tmp;
+ }
+
+ /*
+ * fill in 'header' values
+ */
result[0] = AlignmentAnnotation.SEQUENCE_PROFILE;
- System.arraycopy(rtnval, 0, result, 1, rtnval.length);
+ result[1] = nonZeroCount;
+ result[2] = totalPercentage;
return result;
}
* contains
*
* <pre>
- * [profileType, numberOfValues, totalCount, charValue1, percentage1, charValue2, percentage2, ...]
+ * [profileType, numberOfValues, totalPercentage, charValue1, percentage1, charValue2, percentage2, ...]
* in descending order of percentage value, where the character values encode codon triplets
* </pre>
*
* @param hashtable
* @return
*/
- public static int[] extractCdnaProfile(Hashtable hashtable,
- boolean ignoreGaps)
+ public static int[] extractCdnaProfile(
+ Hashtable<String, Object> hashtable, boolean ignoreGaps)
{
// this holds #seqs, #ungapped, and then codon count, indexed by encoded
// codon triplet
{
break; // nothing else of interest here
}
+ final int percentage = codonCount * 100 / divisor;
+ if (percentage == 0)
+ {
+ /*
+ * this (and any remaining) values rounded down to 0 - discard
+ */
+ break;
+ }
distinctValuesCount++;
result[j++] = codons[i];
- final int percentage = codonCount * 100 / divisor;
result[j++] = percentage;
totalPercentage += percentage;
}
* the consensus data stores to be populated (one per column)
*/
public static void calculateCdna(AlignmentI alignment,
- Hashtable[] hconsensus)
+ Hashtable<String, Object>[] hconsensus)
{
final char gapCharacter = alignment.getGapCharacter();
List<AlignedCodonFrame> mappings = alignment.getCodonFrames();
for (int col = 0; col < cols; col++)
{
// todo would prefer a Java bean for consensus data
- Hashtable<String, int[]> columnHash = new Hashtable<>();
+ Hashtable<String, Object> columnHash = new Hashtable<>();
// #seqs, #ungapped seqs, counts indexed by (codon encoded + 1)
int[] codonCounts = new int[66];
codonCounts[0] = alignment.getSequences().size();
{
codonCounts[codonEncoded + 2]++;
ungappedCount++;
+ break;
}
}
}
*/
public static void completeCdnaConsensus(
AlignmentAnnotation consensusAnnotation,
- Hashtable[] consensusData, boolean showProfileLogo, int nseqs)
+ Hashtable<String, Object>[] consensusData,
+ boolean showProfileLogo, int nseqs)
{
if (consensusAnnotation == null
|| consensusAnnotation.annotations == null
consensusAnnotation.scaleColLabel = true;
for (int col = 0; col < consensusData.length; col++)
{
- Hashtable hci = consensusData[col];
+ Hashtable<String, Object> hci = consensusData[col];
if (hci == null)
{
// gapped protein column?