/**
* 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;
}
/**
* Predict the effect of a seqChange
*
* @param seqChange : Sequence change
* @param seqChangeRef : Before analyzing results, we have to change markers using seqChangerRef
* to create a new reference 'on the fly'
*/
public ChangeEffects seqChangeEffect(Variant seqChange, Variant seqChangeRef) {
ChangeEffects changeEffects = new ChangeEffects(seqChange, seqChangeRef);
// ---
// Chromosome missing?
// ---
if (Config.get().isErrorOnMissingChromo() && isChromosomeMissing(seqChange)) {
changeEffects.addErrorWarning(ErrorWarningType.ERROR_CHROMOSOME_NOT_FOUND);
return changeEffects;
}
// ---
// Check that this is not a huge deletion.
// Huge deletions would crash the rest of the algorithm, so we need to stop them here.
// ---
if (seqChange.isDel() && (seqChange.size() > HUGE_DELETION_SIZE_THRESHOLD)) {
// Get chromosome
String chromoName = seqChange.getChromosomeName();
Chromosome chr = genome.getChromosome(chromoName);
if (chr.size() > 0) {
double ratio = seqChange.size() / ((double) chr.size());
if (ratio > HUGE_DELETION_RATIO_THRESHOLD) {
changeEffects.add(chr, EffectType.CHROMOSOME_LARGE_DELETION, "");
return changeEffects;
}
}
}
// ---
// Query interval tree: Which intervals does seqChange intersect?
// ---
Markers intersects = query(seqChange);
// Show all results
boolean hitChromo = false, hitSomething = false;
if (intersects.size() > 0) {
for (Marker marker : intersects) {
if (marker instanceof Chromosome) hitChromo = true; // Do we hit any chromosome?
else { // Analyze all markers
marker.seqChangeEffect(seqChange, changeEffects, seqChangeRef);
hitSomething = true;
}
}
}
// Any errors or intergenic (i.e. did not hit any gene)
if (!hitChromo) {
if (Config.get().isErrorChromoHit())
changeEffects.addErrorWarning(ErrorWarningType.ERROR_OUT_OF_CHROMOSOME_RANGE);
} else if (!hitSomething) {
if (Config.get().isOnlyRegulation()) changeEffects.setEffectType(EffectType.NONE);
else changeEffects.setEffectType(EffectType.INTERGENIC);
}
return changeEffects;
}
/**
* 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;
}
/**
* Name of the regions hit by a marker
*
* @param marker
* @param showGeneDetails
* @param compareTemplate
* @param id : Only use genes or transcripts matching this ID
* @return
*/
public Set<String> regions(
Marker marker, boolean showGeneDetails, boolean compareTemplate, String id) {
if (Config.get().isErrorOnMissingChromo() && isChromosomeMissing(marker))
throw new RuntimeEOFException("Chromosome missing for marker: " + marker);
boolean hitChromo = false;
HashSet<String> hits = new HashSet<String>();
Markers intersects = query(marker);
if (intersects.size() > 0) {
for (Marker markerInt : intersects) {
if (markerInt instanceof Chromosome) {
hitChromo = true; // OK (we have to hit a chromosome, otherwise it's an error
hits.add(markerInt.getClass().getSimpleName()); // Add marker name to the list
} else if (markerInt instanceof Gene) {
// Analyze Genes
Gene gene = (Gene) markerInt;
regionsAddHit(hits, gene, marker, showGeneDetails, compareTemplate);
// For all transcripts...
for (Transcript tr : gene) {
if ((id == null)
|| gene.getId().equals(id)
|| tr.getId().equals(id)) { // Mathes ID? (...or no ID to match)
// Does it intersect this transcript?
if (tr.intersects(marker)) {
regionsAddHit(hits, tr, marker, showGeneDetails, compareTemplate);
// Does it intersect a UTR?
for (Utr utr : tr.getUtrs())
if (utr.intersects(marker))
regionsAddHit(hits, utr, marker, showGeneDetails, compareTemplate);
// Does it intersect an exon?
for (Exon ex : tr)
if (ex.intersects(marker))
regionsAddHit(hits, ex, marker, showGeneDetails, compareTemplate);
// Does it intersect an intron?
for (Intron intron : tr.introns())
if (intron.intersects(marker))
regionsAddHit(hits, intron, marker, showGeneDetails, compareTemplate);
}
}
}
} else {
// No ID to match?
if (id == null) regionsAddHit(hits, markerInt, marker, showGeneDetails, compareTemplate);
else {
// Is ID from transcript?
Transcript tr = (Transcript) markerInt.findParent(Transcript.class);
if ((tr != null) && (tr.getId().equals(id))) {
regionsAddHit(
hits,
markerInt,
marker,
showGeneDetails,
compareTemplate); // Transcript ID matches => count
} else {
// Is ID from gene?
Gene gene = (Gene) markerInt.findParent(Gene.class);
if ((gene != null) && (gene.getId().equals(id)))
regionsAddHit(
hits,
markerInt,
marker,
showGeneDetails,
compareTemplate); // Gene ID matches => count
}
}
}
}
}
if (!hitChromo) throw new RuntimeException("ERROR: Out of chromosome range. " + marker);
return hits;
}