public ArrayList<Double> mapPDScoresToAlignment(
ArrayList<Double> pdScores, Sequence seq, ArrayList<Integer> indicesOfRowsWithSumZero) {
ArrayList<Double> result = new ArrayList<Double>();
int seqLength = seq.length();
int nonGapPosition = 0;
Symbol gap = seq.getAlphabet().getGapSymbol();
for (int i = 0; i < seqLength; i++) {
double oneScore = Double.MIN_VALUE;
boolean isSumZero = false;
isSumZero = (indicesOfRowsWithSumZero.contains(i));
boolean isGap = false;
isGap =
(seq.symbolAt(i + 1)
== gap); // note: these +1 in indices are because seq starts from one and not zero!
if (isSumZero) {
oneScore = 0.0;
} else {
oneScore = pdScores.get(nonGapPosition);
nonGapPosition++;
}
result.add(oneScore);
}
return result;
} /*mapPDScoresToAlignment*/
public ArrayList<Double> mapViterbiPathToAlignment(
ArrayList<String> viterbiPath, Sequence seq, ArrayList<Integer> indicesOfRowsWithSumZero) {
ArrayList<Double> result = new ArrayList<Double>();
int seqLength = seq.length();
int nonGapPosition = 0;
Symbol gap = seq.getAlphabet().getGapSymbol();
for (int i = 0; i < seqLength; i++) {
double oneScore = Double.MIN_VALUE;
boolean isSumZero = false;
isSumZero = (indicesOfRowsWithSumZero.contains(i));
boolean isGap = false;
isGap =
(seq.symbolAt(i + 1)
== gap); // note: these +1 in indices are because seq and viterbi path start from one
// and not zero!
if (isSumZero) {
oneScore = 0.0;
} else {
if (viterbiPath.get(nonGapPosition).equals("M")) {
oneScore = 0.0;
} else if (viterbiPath.get(nonGapPosition).equals("R")) {
oneScore = 2.0;
} else if (viterbiPath.get(nonGapPosition).equals("r")) {
oneScore = 2.0;
} else if (viterbiPath.get(nonGapPosition).equals("G")) {
oneScore = 2.0;
} else if (viterbiPath.get(nonGapPosition).equals("g")) {
oneScore = 2.0;
} else if (viterbiPath.get(nonGapPosition).equals("E")) {
oneScore = 1.0;
} else if (viterbiPath.get(nonGapPosition).equals("e")) {
oneScore = 1.0;
} else if (viterbiPath.get(nonGapPosition).equals("J")) {
oneScore = 1.5;
} else {
System.err.println("Unknown charecter detected as a state name!");
}
nonGapPosition++;
}
result.add(oneScore);
}
return result;
} /*mapViterbiPathToAlignment*/
/** @return */
public float getPercentGapsQuery() {
return nGapsQ / (float) query.length() * 100;
}
public LinkedHashMap<SimpleAlphabet, SimpleSymbolList> getAlphabetAndSimpleSymbolList(
Matrix2D m, Sequence sequence) throws IllegalSymbolException {
LinkedHashMap<SimpleAlphabet, SimpleSymbolList> alphabetAndSymbolList =
new LinkedHashMap<SimpleAlphabet, SimpleSymbolList>();
SimpleAlphabet alphabet = new SimpleAlphabet();
List<AtomicSymbol> listOfSymbols = new ArrayList<AtomicSymbol>();
alphabet.setName("ObservedSequenceAlphabet");
int numberofRows = m.rows();
int seqLength = sequence.length();
if (numberofRows != seqLength) {
System.err.print(
"It was assumed your sequence has a length equal to the number of rows of the matrix, but found a case that is not true!");
}
Symbol gap = sequence.getAlphabet().getGapSymbol();
for (int i = 0; i < numberofRows; i++) {
List<Symbol> oneListOfSymbol = new ArrayList<Symbol>(3);
// red is match, green is flanking and blue is background
double redValue = m.get(i, 0);
double greenValue = m.get(i, 1);
double blueValue = m.get(i, 2);
double onesum = redValue + greenValue + blueValue;
boolean isGap = false;
isGap = (sequence.symbolAt(i + 1) == gap);
if (onesum
== 0) { // sum of this value is supposed to be one, but for some rows it sums up to
// zero, this is to ignore those up to time we found out why these naouthy rows
// sums up to zero!
// continue;
// note these three lines is only a dummy solution for positions where red, green and blue
// are summed up to zero!
redValue = 0.3333;
greenValue = 0.3333;
blueValue = 1 - (redValue + greenValue);
}
// make one triplet symbol from three symbols
Symbol redSymbol = AlphabetManager.createSymbol(Double.toString(redValue));
Symbol greenSymol = AlphabetManager.createSymbol(Double.toString(greenValue));
Symbol blueSymbol = AlphabetManager.createSymbol(Double.toString(blueValue));
oneListOfSymbol.add(redSymbol);
oneListOfSymbol.add(greenSymol);
oneListOfSymbol.add(blueSymbol);
// now create symbol and add it to alphabet
AtomicSymbol oneSym =
(AtomicSymbol)
AlphabetManager.createSymbol(
Annotation.EMPTY_ANNOTATION, oneListOfSymbol, alphabet);
alphabet.addSymbol(oneSym);
listOfSymbols.add(oneSym);
}
SimpleSymbolList ssl = new SimpleSymbolList(alphabet, listOfSymbols);
alphabetAndSymbolList.put(alphabet, ssl);
return alphabetAndSymbolList;
} /*getAlphabetAndSimpleSymbolList*/
/**
* This method provides a BLAST-like formated alignment from the given <code>String</code>s, in
* which the sequence coordinates and the information "Query" or "Sbjct", respectively is added to
* each line. Each line contains <code>width</code> sequence characters including the gap symbols
* plus the meta information. There is one white line between two pairs of sequences.
*
* @param width the number of symbols to be displayed per line.
* @return formated String.
* @throws BioException
*/
public String formatOutput(int width) throws BioException {
int i, j;
/*
* Highlights equal symbols within the alignment, String match/missmatch
* representation
*/
StringBuilder path = new StringBuilder();
for (i = 0; i < Math.min(queryEnd - queryStart, subjectEnd - subjectStart) + 1; i++) {
Symbol a = query.symbolAt(i + queryStart);
Symbol b = subject.symbolAt(i + subjectStart);
if (!a.equals(query.getAlphabet().getGapSymbol())
&& !b.equals(subject.getAlphabet().getGapSymbol())
&& ((subMatrix.getValueAt(a, b) >= 0) || a.equals(b))) {
path.append('|');
} else {
path.append(' ');
}
}
int maxLength = path.length();
/*
* Math.max(queryEnd - queryStart, subjectEnd - subjectStart) + 1;
*/
Formatter output = new Formatter();
output.format("%n Time (ms): %s%n", time);
output.format(" Length: %d%n", maxLength);
output.format(" Score: %d%n", getScore());
output.format(" Query: %s, Length: %d%n", query.getName(), query.length() - nGapsQ);
output.format(" Sbjct: %s, Length: %d%n", subject.getName(), subject.length() - nGapsS);
output.format(
" Identities: %d/%d, i.e., %d %% (query) and %d %% (sbjct)%n",
identicals,
maxLength,
Math.round(getPercentIdentityQuery()),
Math.round(getPercentIdentitySubject()));
output.format(
" Similars: %d/%d, i.e., %d %% (query) and %d %% (sbjct)%n",
similars,
maxLength,
Math.round(getPercentSimilarityQuery()),
Math.round(getPercentSimilaritySubject()));
output.format(
" No. gaps: %d (%d %%) in query and %d (%d %%) in sbjct%n",
nGapsQ, Math.round(getPercentGapsQuery()), nGapsS, Math.round(getPercentGapsTarget()));
int queryLPos = queryStart, queryRPos, pathLPos = 0, pathRPos;
int subjectLPos = subjectStart, subjectRPos;
int ql = queryLPos - 1, qr = queryLPos - 1, qgaps;
int sl = subjectLPos - 1, sr = subjectLPos - 1, sgaps;
int widthLeft = String.valueOf(Math.max(queryStart, queryEnd)).length();
int widthRight = String.valueOf(Math.max(queryEnd, subjectEnd)).length() + 1;
// Take width of the meta information into account.
width = Math.max(width - widthLeft - widthRight - 12, 2);
for (i = 1; i <= Math.ceil((double) maxLength / width); i++) {
// Query
queryRPos =
Math.min(
queryStart + i * width - 1,
Math.min(queryEnd, subjectEnd - subjectStart + queryStart));
qgaps = 0;
for (j = queryLPos; j <= queryRPos; j++) {
if (!query.symbolAt(j).equals(query.getAlphabet().getGapSymbol())) {
qr++;
} else {
qgaps++;
}
}
if (qgaps <= queryRPos - queryLPos) {
ql++;
}
output.format("%nQuery: %" + widthLeft + "d ", ql);
output.format("%s ", query.subStr(queryLPos, queryRPos));
output.format("%-" + widthRight + "d%n", qr);
queryLPos = queryRPos + 1;
ql = qr;
// Path
pathRPos = Math.min(i * width, path.length());
output.format(
"%-" + (widthLeft + 10) + "c%s",
Character.valueOf(' '),
path.substring(pathLPos, pathRPos));
pathLPos = pathRPos;
// Sbjct
subjectRPos =
Math.min(
subjectStart + i * width - 1,
Math.min(queryEnd - queryStart + subjectStart, subjectEnd));
sgaps = 0;
for (j = subjectLPos; j <= subjectRPos; j++) {
if (!subject.symbolAt(j).equals(subject.getAlphabet().getGapSymbol())) {
sr++;
} else {
sgaps++;
}
}
if (sgaps <= subjectRPos - subjectLPos) {
sl++;
}
output.format("%nSbjct: %" + widthLeft + "d ", sl);
output.format("%s ", subject.subStr(subjectLPos, subjectRPos));
output.format("%-" + widthRight + "d%n", sr);
subjectLPos = subjectRPos + 1;
sl = sr;
}
return output.toString();
}
/** @return */
public float getPercentGapsTarget() {
return nGapsS / (float) subject.length() * 100;
}
/** @return */
public int getQueryLength() {
return query.length();
}
/** @return */
public int getSubjectLength() {
return subject.length();
}
/** @return */
public float getPercentSimilarityQuery() {
return similars / (float) query.length() * 100;
}
/** @return */
public float getPercentSimilaritySubject() {
return similars / (float) subject.length() * 100;
}
/** @return */
public float getPercentIdentitySubject() {
return identicals / (float) (subject.length() - nGapsS) * 100;
}
/** @return */
public float getPercentIdentityQuery() {
return identicals / (float) (query.length() - nGapsQ) * 100;
}
/**
* @param query
* @param subject
* @param subMatrix
* @throws IllegalArgumentException
* @throws BioException
*/
public AlignmentPair(Sequence query, Sequence subject, SubstitutionMatrix subMatrix)
throws IllegalArgumentException, BioException {
this(query, subject, 1, query.length(), 1, subject.length(), subMatrix);
}
private void dnaCommand(HttpServletRequest req, DazzleResponse resp, DazzleDataSource dds)
throws IOException, DataSourceException, ServletException, DazzleException {
DazzleReferenceSource drs = (DazzleReferenceSource) dds;
List segments = DazzleTools.getSegments(dds, req, resp);
if (segments.size() == 0) {
throw new DazzleException(
DASStatus.STATUS_BAD_COMMAND_ARGUMENTS, "No segments specified for dna command");
}
// Fetch and validate the requests.
Map segmentResults = new HashMap();
for (Iterator i = segments.iterator(); i.hasNext(); ) {
Segment seg = (Segment) i.next();
try {
Sequence seq = drs.getSequence(seg.getReference());
if (seq.getAlphabet() != DNATools.getDNA()) {
throw new DazzleException(
DASStatus.STATUS_SERVER_ERROR,
"Sequence " + seg.toString() + " is not in the DNA alphabet");
}
if (seg.isBounded()) {
if (seg.getMin() < 1 || seg.getMax() > seq.length()) {
throw new DazzleException(
DASStatus.STATUS_BAD_COORDS,
"Segment " + seg.toString() + " doesn't fit sequence of length " + seq.length());
}
}
segmentResults.put(seg, seq);
} catch (NoSuchElementException ex) {
throw new DazzleException(DASStatus.STATUS_BAD_REFERENCE, ex);
} catch (DataSourceException ex) {
throw new DazzleException(DASStatus.STATUS_SERVER_ERROR, ex);
}
}
//
// Looks okay -- generate the response document
//
XMLWriter xw = resp.startDasXML("DASDNA", "dasdna.dtd");
try {
xw.openTag("DASDNA");
for (Iterator i = segmentResults.entrySet().iterator(); i.hasNext(); ) {
Map.Entry me = (Map.Entry) i.next();
Segment seg = (Segment) me.getKey();
Sequence seq = (Sequence) me.getValue();
xw.openTag("SEQUENCE");
xw.attribute("id", seg.getReference());
xw.attribute("version", drs.getLandmarkVersion(seg.getReference()));
if (seg.isBounded()) {
xw.attribute("start", "" + seg.getStart());
xw.attribute("stop", "" + seg.getStop());
} else {
xw.attribute("start", "" + 1);
xw.attribute("stop", "" + seq.length());
}
SymbolList syms = seq;
if (seg.isBounded()) {
syms = syms.subList(seg.getMin(), seg.getMax());
}
if (seg.isInverted()) {
syms = DNATools.reverseComplement(syms);
}
xw.openTag("DNA");
xw.attribute("length", "" + syms.length());
for (int pos = 1; pos <= syms.length(); pos += 60) {
int maxPos = Math.min(syms.length(), pos + 59);
xw.println(syms.subStr(pos, maxPos));
}
xw.closeTag("DNA");
xw.closeTag("SEQUENCE");
}
xw.closeTag("DASDNA");
xw.close();
} catch (Exception ex) {
throw new DazzleException(ex, "Error writing DNA document");
}
}
public void main(String[] args) throws BioException, IOException {
if (count.length != args.length) {
System.err.println("The number of counts and output file arguments does not match!");
System.exit(1);
}
OutputStream[] outputStreams;
if ((args != null) && (args.length > 0)) {
outputStreams = new OutputStream[args.length];
for (int i = 0; i < args.length; i++) {
outputStreams[i] = new BufferedOutputStream(new FileOutputStream(args[i]));
}
} else {
outputStreams = new OutputStream[] {System.out};
}
RichSequenceIterator seqIterator =
RichSequence.IOTools.readFastaDNA(new BufferedReader(new FileReader(seqFile)), null);
List<Sequence> seqs = new ArrayList<Sequence>();
while (seqIterator.hasNext()) {
seqs.add(seqIterator.nextSequence());
}
List<List<Sequence>> chosenSeqs = new ArrayList<List<Sequence>>();
if (!sampleWithReplacement) {
for (int c : count) {
List<Sequence> cseqs = new ArrayList<Sequence>();
chosenSeqs.add(cseqs);
while (c > 0) {
int randSeqIndex = random.nextInt(seqs.size());
cseqs.add(seqs.remove(randSeqIndex));
c--;
}
}
} else if (sampleWithReplacement || (length > 0)) {
/*
* if you want to sample from sequences with replacement
* or if the wanted length is specified
*/
for (int c : count) {
int i = 0;
List<Sequence> cseqs = new ArrayList<Sequence>();
chosenSeqs.add(cseqs);
while (c > 0) {
Sequence randomSeq = seqs.get(random.nextInt(seqs.size()));
if (length > 0) {
int startPos = random.nextInt(1 + randomSeq.length() - length);
cseqs.add(
new SimpleSequence(
randomSeq.subList(startPos, startPos + length),
null,
randomSeq.getName() + "_" + i++,
Annotation.EMPTY_ANNOTATION));
} else {
cseqs.add(randomSeq);
}
c--;
}
}
}
int i = 0;
for (List<Sequence> seqList : chosenSeqs) {
for (Sequence seq : seqList) {
Sequence s;
if (uniqueNames) {
s =
new SimpleSequence(
seq.subList(1, seq.length()),
null,
seq.getName() + "_" + i,
Annotation.EMPTY_ANNOTATION);
} else {
s = seq;
}
RichSequence.IOTools.writeFasta(outputStreams[i], s, null);
outputStreams[i].flush();
}
i++;
}
seqs = null;
}
