import jalview.datamodel.AlignedCodonFrame.SequenceToSequenceMapping;
import jalview.io.FastaFile;
import jalview.util.Comparison;
+import jalview.util.LinkedIdentityHashSet;
import jalview.util.MessageManager;
import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.BitSet;
import java.util.Collections;
import java.util.Enumeration;
+import java.util.HashMap;
import java.util.HashSet;
import java.util.Hashtable;
+import java.util.Iterator;
import java.util.List;
import java.util.Map;
import java.util.Set;
* @author JimP
*
*/
-public class Alignment implements AlignmentI
+public class Alignment implements AlignmentI, AutoCloseable
{
private Alignment dataset;
- protected List<SequenceI> sequences;
+ private List<SequenceI> sequences;
protected List<SequenceGroup> groups;
protected char gapCharacter = '-';
- protected int type = NUCLEOTIDE;
-
- public static final int PROTEIN = 0;
-
- public static final int NUCLEOTIDE = 1;
+ private boolean nucleotide = true;
public boolean hasRNAStructure = false;
HiddenSequences hiddenSequences;
+ HiddenColumns hiddenCols;
+
public Hashtable alignmentProperties;
private List<AlignedCodonFrame> codonFrameList;
{
groups = Collections.synchronizedList(new ArrayList<SequenceGroup>());
hiddenSequences = new HiddenSequences(this);
- codonFrameList = new ArrayList<AlignedCodonFrame>();
+ hiddenCols = new HiddenColumns();
+ codonFrameList = new ArrayList<>();
- if (Comparison.isNucleotide(seqs))
- {
- type = NUCLEOTIDE;
- }
- else
- {
- type = PROTEIN;
- }
+ nucleotide = Comparison.isNucleotide(seqs);
sequences = Collections.synchronizedList(new ArrayList<SequenceI>());
public Alignment(SeqCigar[] alseqs)
{
SequenceI[] seqs = SeqCigar.createAlignmentSequences(alseqs,
- gapCharacter, new ColumnSelection(), null);
+ gapCharacter, new HiddenColumns(), null);
initAlignment(seqs);
}
*/
public static AlignmentI createAlignment(CigarArray compactAlignment)
{
- throw new Error(
- MessageManager
- .getString("error.alignment_cigararray_not_implemented"));
+ throw new Error(MessageManager
+ .getString("error.alignment_cigararray_not_implemented"));
// this(compactAlignment.refCigars);
}
return AlignmentUtils.getSequencesByName(this);
}
- /**
- * DOCUMENT ME!
- *
- * @param i
- * DOCUMENT ME!
- *
- * @return DOCUMENT ME!
- */
@Override
public SequenceI getSequenceAt(int i)
{
synchronized (sequences)
{
+
if (i > -1 && i < sequences.size())
{
return sequences.get(i);
}
}
+
return null;
}
+ @Override
+ public SequenceI getSequenceAtAbsoluteIndex(int i)
+ {
+ SequenceI seq = null;
+ if (getHiddenSequences().getSize() > 0)
+ {
+ seq = getHiddenSequences().getHiddenSequence(i);
+ if (seq == null)
+ {
+ // didn't find the sequence in the hidden sequences, get it from the
+ // alignment
+ int index = getHiddenSequences().findIndexWithoutHiddenSeqs(i);
+ seq = getSequenceAt(index);
+ }
+ }
+ else
+ {
+ seq = getSequenceAt(i);
+ }
+ return seq;
+ }
+
/**
- * Adds a sequence to the alignment. Recalculates maxLength and size.
+ * Adds a sequence to the alignment. Recalculates maxLength and size. Note
+ * this currently does not recalculate whether or not the alignment is
+ * nucleotide, so mixed alignments may have undefined behaviour.
*
* @param snew
*/
}
@Override
- public void finalize()
+ public void close()
{
if (getDataset() != null)
{
- getDataset().removeAlignmentRef();
+ try
+ {
+ getDataset().removeAlignmentRef();
+ } catch (Throwable e)
+ {
+ e.printStackTrace();
+ }
}
+ nullReferences();
+ }
+
+ /**
+ * Defensively nulls out references in case this object is not garbage
+ * collected
+ */
+ void nullReferences()
+ {
dataset = null;
sequences = null;
groups = null;
}
/**
- * decrement the alignmentRefs counter by one and call finalize if it goes to
- * zero.
+ * decrement the alignmentRefs counter by one and null references if it goes
+ * to zero.
+ *
+ * @throws Throwable
*/
- private void removeAlignmentRef()
+ private void removeAlignmentRef() throws Throwable
{
if (--alignmentRefs == 0)
{
- finalize();
+ nullReferences();
}
}
- /**
- * DOCUMENT ME!
- *
- * @param s
- * DOCUMENT ME!
- */
@Override
public void deleteSequence(SequenceI s)
{
- deleteSequence(findIndex(s));
+ synchronized (sequences)
+ {
+ deleteSequence(findIndex(s));
+ }
}
- /**
- * DOCUMENT ME!
- *
- * @param i
- * DOCUMENT ME!
- */
@Override
public void deleteSequence(int i)
{
- if (i > -1 && i < getHeight())
+ synchronized (sequences)
{
- synchronized (sequences)
+ if (i > -1 && i < getHeight())
{
sequences.remove(i);
hiddenSequences.adjustHeightSequenceDeleted(i);
}
}
+ @Override
+ public void deleteHiddenSequence(int i)
+ {
+ synchronized (sequences)
+ {
+ if (i > -1 && i < getHeight())
+ {
+ sequences.remove(i);
+ }
+ }
+ }
+
/*
* (non-Javadoc)
*
* @see jalview.datamodel.AlignmentI#findGroup(jalview.datamodel.SequenceI)
*/
@Override
- public SequenceGroup findGroup(SequenceI s)
+ public SequenceGroup findGroup(SequenceI seq, int position)
{
synchronized (groups)
{
- for (int i = 0; i < this.groups.size(); i++)
+ for (SequenceGroup sg : groups)
{
- SequenceGroup sg = groups.get(i);
-
- if (sg.getSequences(null).contains(s))
+ if (sg.getSequences(null).contains(seq))
{
- return sg;
+ if (position >= sg.getStartRes() && position <= sg.getEndRes())
+ {
+ return sg;
+ }
}
}
}
@Override
public SequenceGroup[] findAllGroups(SequenceI s)
{
- ArrayList<SequenceGroup> temp = new ArrayList<SequenceGroup>();
+ ArrayList<SequenceGroup> temp = new ArrayList<>();
synchronized (groups)
{
return;
}
}
- sg.setContext(this);
+ sg.setContext(this, true);
groups.add(sg);
}
}
return;
}
// remove annotation very quickly
- AlignmentAnnotation[] t, todelete = new AlignmentAnnotation[annotations.length], tokeep = new AlignmentAnnotation[annotations.length];
+ AlignmentAnnotation[] t,
+ todelete = new AlignmentAnnotation[annotations.length],
+ tokeep = new AlignmentAnnotation[annotations.length];
int i, p, k;
if (gp == null)
{
}
for (SequenceGroup sg : groups)
{
- sg.setContext(null);
+ sg.setContext(null, false);
}
groups.clear();
}
{
removeAnnotationForGroup(g);
groups.remove(g);
- g.setContext(null);
+ g.setContext(null, false);
}
}
}
int i = 0;
SequenceI sq = null;
String sqname = null;
+ int nseq = sequences.size();
if (startAfter != null)
{
// try to find the sequence in the alignment
boolean matched = false;
- while (i < sequences.size())
+ while (i < nseq)
{
if (getSequenceAt(i++) == startAfter)
{
i = 0;
}
}
- while (i < sequences.size())
+ while (i < nseq)
{
sq = getSequenceAt(i);
sqname = sq.getName();
if (sqname.equals(token) // exact match
|| (b && // allow imperfect matches - case varies
- (sqname.equalsIgnoreCase(token))))
+ (sqname.equalsIgnoreCase(token))))
{
return getSequenceAt(i);
}
* jalview.datamodel.AlignmentI#findIndex(jalview.datamodel.SearchResults)
*/
@Override
- public int findIndex(SearchResults results)
+ public int findIndex(SearchResultsI results)
{
int i = 0;
return -1;
}
- /**
- * DOCUMENT ME!
- *
- * @return DOCUMENT ME!
- */
@Override
public int getHeight()
{
return sequences.size();
}
- /**
- * DOCUMENT ME!
- *
- * @return DOCUMENT ME!
- */
+ @Override
+ public int getAbsoluteHeight()
+ {
+ return sequences.size() + getHiddenSequences().getSize();
+ }
+
@Override
public int getWidth()
{
for (int i = 0; i < sequences.size(); i++)
{
- if (getSequenceAt(i).getLength() > maxLength)
- {
- maxLength = getSequenceAt(i).getLength();
- }
+ maxLength = Math.max(maxLength, getSequenceAt(i).getLength());
}
-
return maxLength;
}
+ @Override
+ public int getVisibleWidth()
+ {
+ int w = getWidth();
+ if (hiddenCols != null)
+ {
+ w -= hiddenCols.getSize();
+ }
+ return w;
+ }
+
/**
* DOCUMENT ME!
*
return true;
}
+ @Override
+ public boolean isHidden(int alignmentIndex)
+ {
+ return (getHiddenSequences().getHiddenSequence(alignmentIndex) != null);
+ }
+
/**
* Delete all annotations, including auto-calculated if the flag is set true.
* Returns true if at least one annotation was deleted, else false.
}
@Override
- public void setNucleotide(boolean b)
- {
- if (b)
- {
- type = NUCLEOTIDE;
- }
- else
- {
- type = PROTEIN;
- }
- }
-
- @Override
public boolean isNucleotide()
{
- if (type == NUCLEOTIDE)
- {
- return true;
- }
- else
- {
- return false;
- }
+ return nucleotide;
}
@Override
}
else if (dataset == null && data != null)
{
+ if (data == this)
+ {
+ throw new IllegalArgumentException("Circular dataset reference");
+ }
if (!(data instanceof Alignment))
{
throw new Error(
private void resolveAndAddDatasetSeq(SequenceI currentSeq,
Set<SequenceI> seqs, boolean createDatasetSequence)
{
+ SequenceI alignedSeq = currentSeq;
if (currentSeq.getDatasetSequence() != null)
{
currentSeq = currentSeq.getDatasetSequence();
currentSeq = currentSeq.createDatasetSequence();
}
}
- if (seqs.contains(currentSeq))
- {
- return;
- }
- List<SequenceI> toProcess = new ArrayList<SequenceI>();
+
+ List<SequenceI> toProcess = new ArrayList<>();
toProcess.add(currentSeq);
while (toProcess.size() > 0)
{
// use a queue ?
SequenceI curDs = toProcess.remove(0);
- if (seqs.contains(curDs))
+
+ if (!seqs.add(curDs))
{
continue;
}
- seqs.add(curDs);
// iterate over database references, making sure we add forward referenced
// sequences
if (curDs.getDBRefs() != null)
{
if (dbr.getMap() != null && dbr.getMap().getTo() != null)
{
+ if (dbr.getMap().getTo() == alignedSeq)
+ {
+ /*
+ * update mapping to be to the newly created dataset sequence
+ */
+ dbr.getMap().setTo(currentSeq);
+ }
if (dbr.getMap().getTo().getDatasetSequence() != null)
{
- throw new Error("Implementation error: Map.getTo() for dbref"
- + dbr + " is not a dataset sequence.");
- // TODO: if this happens, could also rewrite the reference to
- // point to new dataset sequence
+ throw new Error("Implementation error: Map.getTo() for dbref "
+ + dbr + " from " + curDs.getName()
+ + " is not a dataset sequence.");
}
// we recurse to add all forward references to dataset sequences via
// DBRefs/etc
return;
}
// try to avoid using SequenceI.equals at this stage, it will be expensive
- Set<SequenceI> seqs = new jalview.util.LinkedIdentityHashSet<SequenceI>();
+ Set<SequenceI> seqs = new LinkedIdentityHashSet<>();
for (int i = 0; i < getHeight(); i++)
{
int maxLength = -1;
SequenceI current;
- for (int i = 0; i < sequences.size(); i++)
+ int nseq = sequences.size();
+ for (int i = 0; i < nseq; i++)
{
current = getSequenceAt(i);
for (int j = current.getLength(); j > maxLength; j--)
maxLength++;
int cLength;
- for (int i = 0; i < sequences.size(); i++)
+ for (int i = 0; i < nseq; i++)
{
current = getSequenceAt(i);
cLength = current.getLength();
current = getSequenceAt(i);
// This should really be a sequence method
ends[i * 2] = current.findIndex(current.getStart());
- ends[i * 2 + 1] = current.findIndex(current.getStart()
- + current.getLength());
+ ends[i * 2 + 1] = current
+ .findIndex(current.getStart() + current.getLength());
boolean hitres = false;
for (int j = 0, rs = 0, ssiz = current.getLength(); j < ssiz; j++)
{
}
@Override
+ public HiddenColumns getHiddenColumns()
+ {
+ return hiddenCols;
+ }
+
+ @Override
public CigarArray getCompactAlignment()
{
synchronized (sequences)
{
return null;
}
- List<AlignedCodonFrame> cframes = new ArrayList<AlignedCodonFrame>();
+ List<AlignedCodonFrame> cframes = new ArrayList<>();
for (AlignedCodonFrame acf : getCodonFrames())
{
if (acf.involvesSequence(seq))
{
// TODO JAL-1270 needs test coverage
// currently tested for use in jalview.gui.SequenceFetcher
- boolean samegap = toappend.getGapCharacter() == getGapCharacter();
char oldc = toappend.getGapCharacter();
+ boolean samegap = oldc == getGapCharacter();
boolean hashidden = toappend.getHiddenSequences() != null
&& toappend.getHiddenSequences().hiddenSequences != null;
// get all sequences including any hidden ones
- List<SequenceI> sqs = (hashidden) ? toappend.getHiddenSequences()
- .getFullAlignment().getSequences() : toappend.getSequences();
+ List<SequenceI> sqs = (hashidden)
+ ? toappend.getHiddenSequences().getFullAlignment()
+ .getSequences()
+ : toappend.getSequences();
if (sqs != null)
{
// avoid self append deadlock by
- List<SequenceI> toappendsq = new ArrayList<SequenceI>();
+ List<SequenceI> toappendsq = new ArrayList<>();
synchronized (sqs)
{
for (SequenceI addedsq : sqs)
{
if (!samegap)
{
- char[] oldseq = addedsq.getSequence();
- for (int c = 0; c < oldseq.length; c++)
- {
- if (oldseq[c] == oldc)
- {
- oldseq[c] = gapCharacter;
- }
- }
+ addedsq.replace(oldc, gapCharacter);
}
toappendsq.add(addedsq);
}
if (ourval instanceof String)
{
// append strings
- this.setProperty(k, ((String) ourval) + "; "
- + ((String) toapprop));
+ this.setProperty(k,
+ ((String) ourval) + "; " + ((String) toapprop));
}
else
{
String calcId, boolean autoCalc, SequenceI seqRef,
SequenceGroup groupRef)
{
- assert (name != null);
if (annotations != null)
{
for (AlignmentAnnotation annot : getAlignmentAnnotation())
AlignmentAnnotation annot = new AlignmentAnnotation(name, name,
new Annotation[1], 0f, 0f, AlignmentAnnotation.BAR_GRAPH);
annot.hasText = false;
- annot.setCalcId(new String(calcId));
+ if (calcId != null)
+ {
+ annot.setCalcId(new String(calcId));
+ }
annot.autoCalculated = autoCalc;
if (seqRef != null)
{
@Override
public Iterable<AlignmentAnnotation> findAnnotation(String calcId)
{
- ArrayList<AlignmentAnnotation> aa = new ArrayList<AlignmentAnnotation>();
- for (AlignmentAnnotation a : getAlignmentAnnotation())
+ AlignmentAnnotation[] alignmentAnnotation = getAlignmentAnnotation();
+ if (alignmentAnnotation != null)
{
- if (a.getCalcId() == calcId
- || (a.getCalcId() != null && calcId != null && a.getCalcId()
- .equals(calcId)))
- {
- aa.add(a);
- }
+ return AlignmentAnnotation.findAnnotation(
+ Arrays.asList(getAlignmentAnnotation()), calcId);
}
- return aa;
+ return Arrays.asList(new AlignmentAnnotation[] {});
}
- /**
- * Returns an iterable collection of any annotations that match on given
- * sequence ref, calcId and label (ignoring null values).
- */
@Override
public Iterable<AlignmentAnnotation> findAnnotations(SequenceI seq,
String calcId, String label)
{
- ArrayList<AlignmentAnnotation> aa = new ArrayList<AlignmentAnnotation>();
- for (AlignmentAnnotation ann : getAlignmentAnnotation())
- {
- if (ann.getCalcId() != null && ann.getCalcId().equals(calcId)
- && ann.sequenceRef != null && ann.sequenceRef == seq
- && ann.label != null && ann.label.equals(label))
- {
- aa.add(ann);
- }
- }
- return aa;
+ return AlignmentAnnotation.findAnnotations(
+ Arrays.asList(getAlignmentAnnotation()), seq, calcId, label);
}
@Override
@Override
public String toString()
{
- return new FastaFile().print(getSequencesArray());
+ return new FastaFile().print(getSequencesArray(), true);
}
/**
@Override
public Set<String> getSequenceNames()
{
- Set<String> names = new HashSet<String>();
+ Set<String> names = new HashSet<>();
for (SequenceI seq : getSequences())
{
names.add(seq.getName());
}
@Override
- public int[] getVisibleStartAndEndIndex(List<int[]> hiddenCols)
+ public boolean setHiddenColumns(HiddenColumns cols)
+ {
+ boolean changed = cols == null ? hiddenCols != null
+ : !cols.equals(hiddenCols);
+ hiddenCols = cols;
+ return changed;
+ }
+ @Override
+ public void setupJPredAlignment()
+ {
+ SequenceI repseq = getSequenceAt(0);
+ setSeqrep(repseq);
+ HiddenColumns cs = new HiddenColumns();
+ cs.hideList(repseq.getInsertions());
+ setHiddenColumns(cs);
+ }
+
+ @Override
+ public HiddenColumns propagateInsertions(SequenceI profileseq,
+ AlignmentView input)
{
- int[] alignmentStartEnd = new int[] { 0, getWidth() - 1 };
- int startPos = alignmentStartEnd[0];
- int endPos = alignmentStartEnd[1];
+ int profsqpos = 0;
+
+ char gc = getGapCharacter();
+ Object[] alandhidden = input.getAlignmentAndHiddenColumns(gc);
+ HiddenColumns nview = (HiddenColumns) alandhidden[1];
+ SequenceI origseq = ((SequenceI[]) alandhidden[0])[profsqpos];
+ return propagateInsertions(profileseq, origseq, nview);
+ }
+
+ /**
+ *
+ * @param profileseq
+ * sequence in al which corresponds to origseq
+ * @param al
+ * alignment which is to have gaps inserted into it
+ * @param origseq
+ * sequence corresponding to profileseq which defines gap map for
+ * modifying al
+ */
+ private HiddenColumns propagateInsertions(SequenceI profileseq,
+ SequenceI origseq, HiddenColumns hc)
+ {
+ // take the set of hidden columns, and the set of gaps in origseq,
+ // and remove all the hidden gaps from hiddenColumns
+
+ // first get the gaps as a Bitset
+ // then calculate hidden ^ not(gap)
+ BitSet gaps = origseq.gapBitset();
+ hc.andNot(gaps);
+
+ // for each sequence in the alignment, except the profile sequence,
+ // insert gaps corresponding to each hidden region but where each hidden
+ // column region is shifted backwards by the number of preceding visible
+ // gaps update hidden columns at the same time
+ HiddenColumns newhidden = new HiddenColumns();
+
+ int numGapsBefore = 0;
+ int gapPosition = 0;
+ Iterator<int[]> it = hc.iterator();
+ while (it.hasNext())
+ {
+ int[] region = it.next();
+
+ // get region coordinates accounting for gaps
+ // we can rely on gaps not being *in* hidden regions because we already
+ // removed those
+ while (gapPosition < region[0])
+ {
+ gapPosition++;
+ if (gaps.get(gapPosition))
+ {
+ numGapsBefore++;
+ }
+ }
- int[] lowestRange = new int[] { -1, -1 };
- int[] higestRange = new int[] { -1, -1 };
+ int left = region[0] - numGapsBefore;
+ int right = region[1] - numGapsBefore;
- for (int[] hiddenCol : hiddenCols)
+ newhidden.hideColumns(left, right);
+ padGaps(left, right, profileseq);
+ }
+ return newhidden;
+ }
+
+ /**
+ * Pad gaps in all sequences in alignment except profileseq
+ *
+ * @param left
+ * position of first gap to insert
+ * @param right
+ * position of last gap to insert
+ * @param profileseq
+ * sequence not to pad
+ */
+ private void padGaps(int left, int right, SequenceI profileseq)
+ {
+ char gc = getGapCharacter();
+
+ // make a string with number of gaps = length of hidden region
+ StringBuilder sb = new StringBuilder();
+ for (int g = 0; g < right - left + 1; g++)
{
- lowestRange = (hiddenCol[0] <= startPos) ? hiddenCol : lowestRange;
- higestRange = (hiddenCol[1] >= endPos) ? hiddenCol : higestRange;
+ sb.append(gc);
}
- if (lowestRange[0] == -1 && lowestRange[1] == -1)
+ // loop over the sequences and pad with gaps where required
+ for (int s = 0, ns = getHeight(); s < ns; s++)
{
- startPos = alignmentStartEnd[0];
+ SequenceI sqobj = getSequenceAt(s);
+ if ((sqobj != profileseq) && (sqobj.getLength() >= left))
+ {
+ String sq = sqobj.getSequenceAsString();
+ sqobj.setSequence(
+ sq.substring(0, left) + sb.toString() + sq.substring(left));
+ }
}
- else
+ }
+
+ Map<Object, ContactMatrixI> contactmaps = new HashMap<>();
+ @Override
+ public
+ ContactListI getContactListFor(AlignmentAnnotation _aa, int column)
+ {
+ ContactMatrixI cm = contactmaps.get(_aa.annotationId);
+ if (cm == null)
{
- startPos = lowestRange[1] + 1;
+ return null;
}
+ return cm.getContactList(column);
+ }
- if (higestRange[0] == -1 && higestRange[1] == -1)
+ @Override
+ public AlignmentAnnotation addContactList(ContactMatrixI cm)
+ {
+ Annotation _aa[] = new Annotation[getWidth()];
+ Annotation dummy = new Annotation(0.0f);
+ for (int i = 0; i < _aa.length; _aa[i++] = dummy)
{
- endPos = alignmentStartEnd[1];
+ ;
}
- else
+ AlignmentAnnotation aa = new AlignmentAnnotation("Contact Matrix",
+ "Contact Matrix", _aa);
+ aa.graph = AlignmentAnnotation.CUSTOMRENDERER;
+ aa.graphMin = cm.getMin();
+ aa.graphMax = cm.getMax();
+ aa.editable = false;
+ // aa.autoCalculated = true;
+ contactmaps.put(aa.annotationId, cm);
+ // TODO: contact matrices could be intra or inter - more than one refseq possible!
+ if (cm.hasReferenceSeq())
{
- endPos = higestRange[0] - 1;
+ aa.setSequenceRef(cm.getReferenceSeq());
}
- return new int[] { startPos, endPos };
+ addAnnotation(aa);
+ return aa;
}
}
git://source.jalview.org / jalview.git / blobdiff