/**
* Count number of bases, for a given chromosome and marker type
*
* @param mtype
* @param chr
* @param markers
* @return
*/
void countBases(String mtype, Chromosome chr, Markers markers) {
String chrName = chr.getChromosomeName();
if (verbose) System.err.print(" " + chrName);
// Initialize
byte busy[] = new byte[chr.size()];
for (int i = 0; i < busy.length; i++) busy[i] = 0;
for (Marker m : markers) {
// Same marker type & same chromo? Count bases
if (m.getChromosomeName().equals(chrName) && markerTypes.isType(m, mtype)) {
for (int i = m.getStart(); i <= m.getEnd(); i++) busy[i] = 1;
}
}
int latest = 0;
for (int i = 0; i < busy.length; i++) {
// Transition? Count another marker
if ((i > 0) && (busy[i] != 0) && (busy[i - 1] == 0)) {
if ((i - latest) <= readLength)
countBases.inc(mtype, i - latest); // Intervals are less than one read away? Unify them
else countMarkers.inc(mtype);
}
// Base busy? Count another base
if (busy[i] != 0) {
countBases.inc(mtype);
latest = i;
}
}
}
/**
* 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;
}
/**
* Is the chromosome missing in this marker?
*
* @param marker
* @return
*/
boolean isChromosomeMissing(Marker marker) {
// Missing chromosome in marker?
if (marker.getChromosome() == null) return true;
// Missing chromosome in genome?
String chrName = marker.getChromosomeName();
Chromosome chr = genome.getChromosome(chrName);
if (chr == null) return true;
// Chromosome length is 1 or less?
if (chr.size() < 1) return true;
// Tree not found in interval forest?
if (!intervalForest.hasTree(chrName)) return true;
// OK, we have the chromosome
return false;
}