/** Parse genotype string (sparse matrix) and set all entries using 'value' */
void parseSparseGt(String str, byte gt[], int valueInt) {
if ((str == null) || (str.isEmpty()) || (str.equals("true"))) return;
// Split comma separated indeces
String idxs[] = str.split(",");
byte value = (byte) valueInt;
// Set all entries
for (String idx : idxs) {
int i = Gpr.parseIntSafe(idx);
gt[i] = value;
}
}
/** Read and parse genes file */
protected void readRefSeqFile() {
try {
int count = 0;
BufferedReader reader = Gpr.reader(fileName);
if (reader == null) return; // Error
for (lineNum = 1; reader.ready(); lineNum++) {
line = reader.readLine();
// Skip headers
if (!line.startsWith("#")) {
String fields[] = line.split("\t");
if (fields.length >= 9) {
// Parse fields
int fieldNum = 0;
String id = fields[fieldNum++];
String chromoName = fields[fieldNum++];
boolean strandMinus = fields[fieldNum++].equals("-");
int txstart = parsePosition(fields[fieldNum++]);
int txend =
parsePosition(fields[fieldNum++])
- 1; // Our internal database representations of coordinates always have a
// zero-based start and a one-based end (Reference:
// http://genome.ucsc.edu/FAQ/FAQtracks.html#tracks1 )
int cdsStart = parsePosition(fields[fieldNum++]);
int cdsEnd =
parsePosition(fields[fieldNum++])
- 1; // Our internal database representations of coordinates always have a
// zero-based start and a one-based end (Reference:
// http://genome.ucsc.edu/FAQ/FAQtracks.html#tracks1 )
int exonCount = Gpr.parseIntSafe(fields[fieldNum++]);
String exonStarts = fields[fieldNum++];
String exonEnds =
fields[
fieldNum++]; // Our internal database representations of coordinates always have
// a zero-based start and a one-based end (Reference:
// http://genome.ucsc.edu/FAQ/FAQtracks.html#tracks1 )
String proteinId = fields[fieldNum++];
// String alignId = fields[fieldNum++]; // Not used
// ---
// Create
// ----
Chromosome chromo = getOrCreateChromosome(chromoName);
// Is it protein coding?
boolean isCoding = !proteinId.isEmpty(); // Protein ID assigned?
// Create IDs
String trId = uniqueTrId(id);
// Get or create gene
Gene gene =
findOrCreateGene(proteinId, trId, chromo, txstart, txend, strandMinus, isCoding);
// Create transcript
Transcript tr = new Transcript(gene, txstart, txend, strandMinus, trId);
tr.setProteinCoding(isCoding);
add(tr);
// Add Exons and CDS
String exStartStr[] = exonStarts.split(",");
String exEndStr[] = exonEnds.split(",");
for (int i = 0; i < exonCount; i++) {
// Exons
int exStart = parsePosition(exStartStr[i]);
int exEnd =
parsePosition(exEndStr[i])
- 1; // Our internal database representations of coordinates always have a
// zero-based start and a one-based end (Reference:
// http://genome.ucsc.edu/FAQ/FAQtracks.html#tracks1 )
String exId = trId + ".ex." + (i + 1);
Exon ex = new Exon(tr, exStart, exEnd, strandMinus, exId, i);
add(ex);
// CDS (ony if intersects)
if ((exStart <= cdsEnd) && (exEnd >= cdsStart)) {
Cds cds =
new Cds(
tr,
Math.max(cdsStart, exStart),
Math.min(cdsEnd, exEnd),
strandMinus,
exId);
add(cds);
}
}
count++;
if (count % MARK == 0) System.out.print('.');
if (count % (100 * MARK) == 0) System.out.print("\n\t");
}
}
}
reader.close();
} catch (Exception e) {
Gpr.debug("Offending line (lineNum: " + lineNum + "): '" + line + "'");
throw new RuntimeException(e);
}
}
/** Create a variant */
List<Variant> variants(Chromosome chromo, int start, String reference, String alt, String id) {
List<Variant> list = null;
if (alt != null) alt = alt.toUpperCase();
if (alt == null || alt.isEmpty() || alt.equals(reference)) {
// Non-variant
list = Variant.factory(chromo, start, reference, null, id, false);
} else if (alt.charAt(0) == '<') {
// Structural variants
if (alt.startsWith("<DEL")) {
// Case: Deletion
// 2 321682 . T <DEL> 6 PASS
// IMPRECISE;SVTYPE=DEL;END=321887;SVLEN=-105;CIPOS=-56,20;CIEND=-10,62
String ch = ref;
int startNew = start;
if (end > start) {
startNew = start + reference.length();
int size = end - startNew + 1;
char change[] = new char[size];
for (int i = 0; i < change.length; i++)
change[i] = reference.length() > i ? reference.charAt(i) : 'N';
ch = new String(change);
}
list = Variant.factory(chromo, startNew, ch, "", id, false);
} else if (alt.startsWith("<INV")) {
// Inversion
int startNew = start + reference.length();
Variant var = new Variant(chromo, startNew, end, id);
var.setVariantType(VariantType.INV);
list = new LinkedList<>();
list.add(var);
} else if (alt.startsWith("<DUP")) {
// Duplication
int startNew = start + reference.length();
Variant var = new Variant(chromo, startNew, end, id);
var.setVariantType(VariantType.DUP);
list = new LinkedList<>();
list.add(var);
}
} else if ((alt.indexOf('[') >= 0) || (alt.indexOf(']') >= 0)) {
// Translocations
// Parse ALT string
boolean left = alt.indexOf(']') >= 0;
String sep = (left ? "\\]" : "\\[");
String tpos[] = alt.split(sep);
String pos = tpos[1];
boolean before = (alt.indexOf(']') == 0) || (alt.indexOf('[') == 0);
String altBases = (before ? tpos[2] : tpos[0]);
// Parse 'chr:start'
String posSplit[] = pos.split(":");
String trChrName = posSplit[0];
Chromosome trChr = chromo.getGenome().getOrCreateChromosome(trChrName);
int trStart = Gpr.parseIntSafe(posSplit[1]) - 1;
VariantBnd var = new VariantBnd(chromo, start, ref, altBases, trChr, trStart, left, before);
list = new LinkedList<>();
list.add(var);
} else if (reference.length() == alt.length()) {
// Case: SNP, MNP
if (reference.length() == 1) {
// SNPs
// 20 3 . C G . PASS DP=100
list = Variant.factory(chromo, start, reference, alt, id, true);
} else {
// MNPs
// 20 3 . TC AT . PASS DP=100
// Sometimes the first bases are the same and we can trim them
int startDiff = Integer.MAX_VALUE;
for (int i = 0; i < reference.length(); i++)
if (reference.charAt(i) != alt.charAt(i)) startDiff = Math.min(startDiff, i);
// MNPs
// Sometimes the last bases are the same and we can trim them
int endDiff = 0;
for (int i = reference.length() - 1; i >= 0; i--)
if (reference.charAt(i) != alt.charAt(i)) endDiff = Math.max(endDiff, i);
String newRef = reference.substring(startDiff, endDiff + 1);
String newAlt = alt.substring(startDiff, endDiff + 1);
list = Variant.factory(chromo, start + startDiff, newRef, newAlt, id, true);
}
} else {
// Short Insertions, Deletions or Mixed Variants (substitutions)
VcfRefAltAlign align = new VcfRefAltAlign(alt, reference);
align.align();
int startDiff = align.getOffset();
switch (align.getVariantType()) {
case DEL:
// Case: Deletion
// 20 2 . TC T . PASS DP=100
// 20 2 . AGAC AAC . PASS DP=100
String ref = "";
String ch = align.getAlignment();
if (!ch.startsWith("-"))
throw new RuntimeException(
"Deletion '" + ch + "' does not start with '-'. This should never happen!");
list = Variant.factory(chromo, start + startDiff, ref, ch, id, true);
break;
case INS:
// Case: Insertion of A { tC ; tCA } tC is the reference allele
// 20 2 . TC TCA . PASS DP=100
ch = align.getAlignment();
ref = "";
if (!ch.startsWith("+"))
throw new RuntimeException(
"Insertion '" + ch + "' does not start with '+'. This should never happen!");
list = Variant.factory(chromo, start + startDiff, ref, ch, id, true);
break;
case MIXED:
// Case: Mixed variant (substitution)
reference = reference.substring(startDiff);
alt = alt.substring(startDiff);
list = Variant.factory(chromo, start + startDiff, reference, alt, id, true);
break;
default:
// Other change type?
throw new RuntimeException(
"Unsupported VCF change type '"
+ align.getVariantType()
+ "'\n\tRef: "
+ reference
+ "'\n\tAlt: '"
+ alt
+ "'\n\tVcfEntry: "
+ this);
}
}
// ---
// Add original 'ALT' field as genotype
// ---
if (list == null) list = new LinkedList<>();
for (Variant variant : list) variant.setGenotype(alt);
return list;
}