/** Count bases covered for each marker type */
public void countBases() {
// ---
// Add all markers
// ---
Markers markers = new Markers();
markers.add(snpEffectPredictor.getMarkers());
for (Gene gene : snpEffectPredictor.getGenome().getGenes()) {
markers.add(gene);
markers.add(gene.markers());
}
for (Chromosome chr : snpEffectPredictor.getGenome()) markers.add(chr);
// ---
// Calculate raw counts
// ---
for (Marker m : markers) {
String mtype = markerTypes.getType(m);
String msubtype = markerTypes.getSubType(m);
rawCountMarkers.inc(mtype);
rawCountBases.inc(mtype, m.size());
// Count sub-types (if any)
if (msubtype != null) {
rawCountMarkers.inc(msubtype);
rawCountBases.inc(msubtype, m.size());
}
}
// ---
// Count number of bases for each marker type (overlap and join)
// ---
for (String mtype : rawCountMarkers.keysSorted()) {
if (mtype.equals(Chromosome.class.getSimpleName()))
continue; // We calculate chromosomes later (it's faster)
if (verbose) System.err.print(mtype + ":");
if (countMarkers.get(mtype) == 0) {
for (Chromosome chr : snpEffectPredictor.getGenome()) countBases(mtype, chr, markers);
}
if (verbose) System.err.println("");
}
// Show chromosomes length
String mtype = Chromosome.class.getSimpleName();
for (Chromosome chr : snpEffectPredictor.getGenome()) {
countBases.inc(mtype, chr.size());
countMarkers.inc(mtype);
}
}
/**
* Return a collection of intervals that intersect 'marker' Query resulting genes, transcripts and
* exons to get ALL types of intervals possible
*
* @return
*/
public Markers queryDeep(Marker marker) {
if (Config.get().isErrorOnMissingChromo() && isChromosomeMissing(marker))
throw new RuntimeEOFException("Chromosome missing for marker: " + marker);
boolean hitChromo = false;
Markers hits = new Markers();
Markers intersects = query(marker);
if (intersects.size() > 0) {
for (Marker m : intersects) {
hits.add(m);
if (m instanceof Chromosome) {
hitChromo = true; // OK (we have to hit a chromosome, otherwise it's an error
} else if (m instanceof Gene) {
// Analyze Genes
Gene gene = (Gene) m;
hits.addAll(gene.query(marker));
}
}
}
if (!hitChromo && Config.get().isErrorChromoHit())
throw new RuntimeException("ERROR: Out of chromosome range. " + marker);
return hits;
}
/** Save nextprot markers */
void save() {
String nextProtBinFile = config.getDirDataVersion() + "/nextProt.bin";
if (verbose) Timer.showStdErr("Saving database to file '" + nextProtBinFile + "'");
// Add chromosomes
HashSet<Chromosome> chromos = new HashSet<Chromosome>();
for (Marker m : markers) chromos.add(m.getChromosome());
// Create a set of all markers to be saved
Markers markersToSave = new Markers();
markersToSave.add(genome);
for (Chromosome chr : chromos) markersToSave.add(chr);
for (Marker m : markers) markersToSave.add(m);
// Save
MarkerSerializer markerSerializer = new MarkerSerializer();
markerSerializer.save(nextProtBinFile, markersToSave);
}
/**
* Create (and add) up-down stream, splice sites, intergenic, etc
*
* @return
*/
public Markers createGenomicRegions() {
Markers markers = new Markers();
// Add up-down stream intervals
for (Marker upDownStream : genome.getGenes().createUpDownStream(upDownStreamLength))
markers.add(upDownStream);
// Add splice site intervals
for (Marker spliceSite :
genome
.getGenes()
.createSpliceSites(
spliceSiteSize, spliceRegionExonSize, spliceRegionIntronMin, spliceRegionIntronMax))
markers.add(spliceSite);
// Intergenic markers
for (Intergenic intergenic : genome.getGenes().createIntergenic()) markers.add(intergenic);
return markers;
}
/**
* Find closest gene to this marker
*
* <p>In case more than one 'closest' gene is found (e.g. two or more genes at the same distance).
* The following rules apply:
*
* <p>i) If many genes have the same 'closest distance', coding genes are preferred.
*
* <p>ii) If more than one coding gene has the same 'closet distance', a random gene is returned.
*
* @param inputInterval
*/
public Gene queryClosestGene(Marker inputInterval) {
int initialExtension = 1000;
String chrName = inputInterval.getChromosomeName();
Chromosome chr = genome.getChromosome(chrName);
if (chr == null) return null;
if (chr.size() > 0) {
// Extend interval to capture 'close' genes
for (int extend = initialExtension; extend < chr.size(); extend *= 2) {
int start = Math.max(inputInterval.getStart() - extend, 0);
int end = inputInterval.getEnd() + extend;
Marker extended = new Marker(chr, start, end, 1, "");
// Find all genes that intersect with the interval
Markers markers = query(extended);
Markers genes = new Markers();
int minDist = Integer.MAX_VALUE;
for (Marker m : markers) {
if (m instanceof Gene) {
int dist = m.distance(inputInterval);
if (dist < minDist) {
genes.add(m);
minDist = dist;
}
}
}
// Found something?
if (genes.size() > 0) {
// Find a gene having distance 'minDist'. Prefer coding genes
Gene minDistGene = null;
for (Marker m : genes) {
int dist = m.distance(inputInterval);
if (dist == minDist) {
Gene gene = (Gene) m;
if (minDistGene == null) minDistGene = gene;
else if (!minDistGene.isProteinCoding() && gene.isProteinCoding()) minDistGene = gene;
}
}
return minDistGene;
}
}
}
// Nothing found
return null;
}
/** Create interval trees (forest) */
public void buildForest() {
intervalForest = new IntervalForest();
// Add all chromosomes to forest
if (useChromosomes) {
for (Chromosome chr : genome) intervalForest.add(chr);
}
// Add all genes to forest
for (Gene gene : genome.getGenes()) intervalForest.add(gene);
// ---
// Create (and add) up-down stream, splice sites, intergenic, etc
// ---
markers.add(createGenomicRegions());
// Add all 'markers' to forest (includes custom intervals)
intervalForest.add(markers);
// Build interval forest
intervalForest.build();
}
/**
* Parse a protein node
*
* @param node
*/
void parseAnnotation(Node ann, String geneId, String category) {
// Description
Node descr = findOneNode(ann, NODE_NAME_DESCRIPTION, null, null, null);
String description = getText(descr);
if (description == null) description = "";
else if (description.indexOf(';') > 0)
description = description.substring(0, description.indexOf(';')); // Cut after semicolon
// Controlled vocabulary
Node cv = findOneNode(ann, NODE_NAME_CVNAME, null, null, null);
String contrVoc = getText(cv);
if (contrVoc == null) contrVoc = "";
contrVoc.indexOf(';');
String cvs[] = contrVoc.split(";", 2);
String contrVoc2 = "";
if (cvs.length > 1) {
contrVoc = cvs[0];
contrVoc2 = cvs[1];
}
// Search annotations
List<Node> posNodes = findNodes(ann, NODE_NAME_POSITION, null, null, null);
for (Node pos : posNodes) {
// Get first & last position
String first = getAttribute(pos, ATTR_NAME_FIRST);
String last = getAttribute(pos, ATTR_NAME_LAST);
int aaStart = Gpr.parseIntSafe(first) - 1;
int aaEnd = Gpr.parseIntSafe(last) - 1;
int len = aaEnd - aaStart + 1;
// Get ID
Node isoAnn = pos.getParentNode().getParentNode();
String isoformRef = getAttribute(isoAnn, ATTR_NAME_ISOFORM_REF);
// Find sequence
String sequence = sequenceByUniqueName.get(isoformRef);
String subSeq = "";
if ((sequence != null) && (aaStart >= 0) && (aaEnd >= aaStart))
subSeq = sequence.substring(aaStart, aaEnd + 1);
// Check transcript
TranscriptData trData = transcriptData(isoformRef, aaStart, aaEnd, sequence, subSeq);
// Create nextProt markers
if (trData.ok && (len > 0)) {
if (debug)
System.out.println(
geneId //
+ "\t"
+ isoformRef //
+ "\t"
+ trData.tr.getId() //
+ "\t"
+ category //
+ "\t"
+ description //
+ "\t"
+ contrVoc //
+ "\t"
+ contrVoc2 //
+ "\t"
+ first //
+ "\t"
+ last //
+ "\t"
+ len //
+ "\t"
+ trData.chrName //
+ "\t"
+ trData.chrPosStart //
+ "\t"
+ trData.chrPosEnd //
+ "\t"
+ subSeq //
+ "\t"
+ trData.codon //
+ "\t"
+ trData.aa //
);
// Create marker
String id = key(category, contrVoc, description);
NextProt nextProt = new NextProt(trData.tr, trData.chrPosStart, trData.chrPosEnd, id);
markers.add(nextProt);
// if (subSeq.length() == 1) countAaSequence(category, contrVoc, description, subSeq);
countAaSequence(category, contrVoc, description, subSeq);
}
}
}
/**
* Add a set of markers
*
* @param markersToAdd
*/
public void addAll(Markers markersToAdd) {
for (Marker marker : markersToAdd) markers.add(marker);
}
/**
* Add a marker
*
* <p>Note: Markers have to be added BEFORE building the interval trees. Interval trees are built
* the first time you call snpEffect(snp) method.
*
* @param marker
*/
public void add(Marker marker) {
markers.add(marker);
}
