/** Parse a 'line' from a 'vcfFileIterator' */
public void parse() {
// Parse line
String fields[] =
line.split("\t", 10); // Only pare the fist 9 fields (i.e. do not parse genotypes)
// Is line OK?
if (fields.length >= 4) {
// Chromosome and position. VCF files are one-base, so inOffset should be 1.
chromosomeName = fields[0].trim();
// Chromosome
Chromosome chromo = vcfFileIterator.getChromosome(chromosomeName);
parent = chromo;
vcfFileIterator.sanityCheckChromo(chromosomeName, chromo); // Sanity check
// Start
start = vcfFileIterator.parsePosition(vcfFileIterator.readField(fields, 1));
// ID (e.g. might indicate dbSnp)
id = vcfFileIterator.readField(fields, 2);
// REF
ref = vcfFileIterator.readField(fields, 3).toUpperCase(); // Reference and change
strandMinus = false; // Strand is always positive (defined in VCF spec.)
// ALT
altStr = vcfFileIterator.readField(fields, 4).toUpperCase();
parseAlts(altStr);
// Quality
String qStr = vcfFileIterator.readField(fields, 5);
if (!qStr.isEmpty()) quality = Gpr.parseDoubleSafe(qStr);
else quality = null;
// Filter
filter = vcfFileIterator.readField(fields, 6); // Filter parameters
// INFO fields
infoStr = vcfFileIterator.readField(fields, 7);
info = null;
// Start & End coordinates are anchored to the reference genome, thus based on REF field (ALT
// is not taken into account)
parseEnd(altStr);
// Genotype format
format = null;
if (fields.length > 8)
format =
vcfFileIterator.readField(
fields, 8); // This field is optional, So it can be null or EMPTY ('.')
// Add genotype fields (lazy parse)
if (fields.length > 9) genotypeFieldsStr = fields[9];
} else
throw new RuntimeException(
"Impropper VCF entry: Not enough fields (missing tab separators?).\n" + line);
}
/**
* Get info field as a 'double' number The norm specifies data type as 'FLOAT', that is why the
* name of this method might be not intuitive
*/
public double getInfoFloat(String key) {
if (info == null) parseInfo();
String f = info.get(key);
if (f == null) return Double.NaN;
return Gpr.parseDoubleSafe(f);
}