protected internal static Genotype Create(string sampleName, List <Allele> alleles, double[] gls)
{
var gb = new GenotypeBuilder(sampleName, alleles);
gb.setPL(gls);
return(gb.Make());
}
public static Genotype Create(string sampleName, List <Allele> alleles, Dictionary <string, object> attributes)
{
var gb = new GenotypeBuilder(sampleName, alleles);
gb.AddAttributes(attributes);
return(gb.Make());
}
private Genotype fullyDecodeGenotypes (Genotype g, VCFHeader header)
{
IDictionary<string, object> map = fullyDecodeAttributes (g.ExtendedAttributes, header, true);
var g2 = new GenotypeBuilder (g);
g2.AddAttributes (map);
return g2.Make ();
}
/// <summary>
/// add the genotype data
/// </summary>
/// <param name="vc"> the variant context </param>
/// <param name="genotypeFormatKeys"> Genotype formatting string </param>
/// <param name="alleleMap"> alleles for this context </param>
/// <exception cref="IOException"> for writer </exception>
private string getGenotypeDataText(VariantContext vc, IDictionary <Allele, string> alleleMap, IList <string> genotypeFormatKeys)
{
StringBuilder sbn = new StringBuilder();
int ploidy = vc.GetMaxPloidy(2);
foreach (string sample in mHeader.GenotypeSampleNames)
{
sbn.Append(VCFConstants.FIELD_SEPARATOR);
Genotype g = vc.GetGenotype(sample);
if (g == null)
{
g = GenotypeBuilder.CreateMissing(sample, ploidy);
}
IList <string> attrs = new List <string>(genotypeFormatKeys.Count);
foreach (string field in genotypeFormatKeys)
{
if (field.Equals(VCFConstants.GENOTYPE_KEY))
{
if (!g.Available)
{
throw new Exception("GTs cannot be missing for some samples if they are available for others in the record");
}
sbn.Append(getAlleleText(g.getAllele(0), alleleMap));
for (int i = 1; i < g.Ploidy; i++)
{
sbn.Append(g.Phased ? VCFConstants.PHASED : VCFConstants.UNPHASED);
sbn.Append(getAlleleText(g.getAllele(i), alleleMap));
}
continue;
}
else
{
string outputValue;
if (field.Equals(VCFConstants.GENOTYPE_FILTER_KEY))
{
outputValue = g.Filtered ? g.Filters : VCFConstants.PASSES_FILTERS_v4;
}
else
{
IntGenotypeFieldAccessors.Accessor accessor = intGenotypeFieldAccessors.GetAccessor(field);
if (accessor != null)
{
int[] intValues = accessor.getValues(g);
if (intValues == null)
{
outputValue = VCFConstants.MISSING_VALUE_v4;
}
else if (intValues.Length == 1) // fast path
{
outputValue = Convert.ToString(intValues[0]);
}
else
{
StringBuilder sb = new StringBuilder();
sb.Append(intValues[0]);
for (int i = 1; i < intValues.Length; i++)
{
sb.Append(",");
sb.Append(intValues[i]);
}
outputValue = sb.ToString();
}
}
else
{
object val = g.HasExtendedAttribute(field) ? g.GetExtendedAttribute(field) : VCFConstants.MISSING_VALUE_v4;
VCFFormatHeaderLine metaData = mHeader.getFormatHeaderLine(field);
if (metaData != null)
{
int numInFormatField = metaData.getCount(vc);
if (numInFormatField > 1 && val.Equals(VCFConstants.MISSING_VALUE_v4))
{
// If we have a missing field but multiple values are expected, we need to construct a new string with all fields.
// For example, if Number=2, the string has to be ".,."
StringBuilder sb = new StringBuilder(VCFConstants.MISSING_VALUE_v4);
for (int i = 1; i < numInFormatField; i++)
{
sb.Append(",");
sb.Append(VCFConstants.MISSING_VALUE_v4);
}
val = sb.ToString();
}
}
// assume that if key is absent, then the given string encoding suffices
outputValue = formatVCFField(val);
}
}
if (outputValue != null)
{
attrs.Add(outputValue);
}
}
}
// strip off trailing missing values
for (int i = attrs.Count - 1; i >= 0; i--)
{
if (isMissingValue(attrs[i]))
{
attrs.RemoveAt(i);
}
else
{
break;
}
}
for (int i = 0; i < attrs.Count; i++)
{
if (i > 0 || genotypeFormatKeys.Contains(VCFConstants.GENOTYPE_KEY))
{
sbn.Append(VCFConstants.GENOTYPE_FIELD_SEPARATOR);
}
sbn.Append(attrs[i]);
}
}
return(sbn.ToString());
}
protected internal static Genotype Create(string sampleName, List<Allele> alleles, double[] gls)
{
var gb = new GenotypeBuilder(sampleName, alleles);
gb.setPL(gls);
return gb.Make();
}
public static Genotype Create(string sampleName, List<Allele> alleles, Dictionary<string, object> attributes)
{
var gb = new GenotypeBuilder(sampleName, alleles);
gb.AddAttributes(attributes);
return gb.Make();
}
/// <summary>
/// Create a genotype map
/// </summary>
/// <param name="str"> the string </param>
/// <param name="alleles"> the list of alleles </param>
/// <returns> a mapping of sample name to genotype object </returns>
public LazyGenotypesContext.LazyData CreateGenotypeMap(string str, IList <Allele> alleles, string chr, int pos)
{
if (genotypeParts == null)
{
genotypeParts = new String[header.ColumnCount - NUM_STANDARD_FIELDS];
}
try {
FastStringUtils.Split(str, VCFConstants.FIELD_SEPARATOR_CHAR, genotypeParts);
} catch (Exception e) {
throw new VCFParsingError("Could not parse genotypes, was expecting " + (genotypeParts.Length - 1).ToString() + " but found " + str.Split(VCFConstants.FIELD_SEPARATOR_CHAR).Length.ToString(), e);
}
List <Genotype> genotypes = new List <Genotype> (genotypeParts.Length);
// get the format keys
//int nGTKeys = ParsingUtils.Split(genotypeParts[0], genotypeKeyArray, VCFConstants.GENOTYPE_FIELD_SEPARATOR_CHAR);
string[] genotypeKeyArray = genotypeParts [0].Split(VCFConstants.GENOTYPE_FIELD_SEPARATOR_CHAR);
int genotypeAlleleLocation = Array.IndexOf(genotypeKeyArray, VCFConstants.GENOTYPE_KEY);
if (version != VCFHeaderVersion.VCF4_1 && genotypeAlleleLocation == -1)
{
generateException("Unable to find the GT field for the record; the GT field is required in VCF4.0");
}
// clear out our allele mapping
alleleMap.Clear();
GenotypeBuilder gb = new GenotypeBuilder();
// cycle through the genotype strings
for (int genotypeOffset = 1; genotypeOffset < genotypeParts.Length; genotypeOffset++)
{
Genotype curGenotype;
string sampleName = header.GenotypeSampleNames [genotypeOffset - 1];
var currentGeno = genotypeParts [genotypeOffset];
//shortcut for null alleles
if (currentGeno == "./.")
{
curGenotype = GenotypeBuilder.CreateMissing(sampleName, 2);
}
else if (currentGeno == ".")
{
curGenotype = GenotypeBuilder.CreateMissing(sampleName, 1);
}
else
{
gb.Reset(false);
gb.SampleName = sampleName;
string[] GTValueArray = FastStringUtils.Split(currentGeno, VCFConstants.GENOTYPE_FIELD_SEPARATOR_CHAR, int.MaxValue, StringSplitOptions.None);
// cycle through the sample names
// check to see if the value list is longer than the key list, which is a problem
if (genotypeKeyArray.Length < GTValueArray.Length)
{
generateException("There are too many keys for the sample " + sampleName + ", line is: keys = " + genotypeParts [0] + ", values = " + genotypeParts [genotypeOffset]);
}
if (genotypeAlleleLocation > 0)
{
generateException("Saw GT field at position " + genotypeAlleleLocation + ", but it must be at the first position for genotypes when present");
}
//TODO: THIS IS A DAMNED MESS
//Code loops over all fields in the key and decodes them, adding them as information to the genotype builder, which then makes it.
if (genotypeKeyArray.Length > 0)
{
gb.MaxAttributes(genotypeKeyArray.Length - 1);
for (int i = 0; i < genotypeKeyArray.Length; i++)
{
string gtKey = genotypeKeyArray [i];
if (i >= GTValueArray.Length)
{
break;
}
// todo -- all of these on the fly parsing of the missing value should be static constants
if (gtKey == VCFConstants.GENOTYPE_FILTER_KEY)
{
IList <string> filters = parseFilters(GetCachedString(GTValueArray [i]));
if (filters != null)
{
gb.SetFilters(filters.ToList());
}
}
else if (GTValueArray [i] == VCFConstants.MISSING_VALUE_v4)
{
// don't add missing values to the map
}
else
{
if (gtKey == VCFConstants.GENOTYPE_QUALITY_KEY)
{
if (GTValueArray [i] == VCFConstants.MISSING_GENOTYPE_QUALITY_v3)
{
gb.noGQ();
}
else
{
gb.GQ = ((int)Math.Round(Convert.ToDouble(GTValueArray [i])));
}
}
else if (gtKey == VCFConstants.GENOTYPE_ALLELE_DEPTHS)
{
gb.AD = (decodeInts(GTValueArray [i]));
}
else if (gtKey == VCFConstants.GENOTYPE_PL_KEY)
{
gb.PL = (decodeInts(GTValueArray [i]));
}
else if (gtKey == VCFConstants.GENOTYPE_LIKELIHOODS_KEY)
{
gb.PL = (GenotypeLikelihoods.fromGLField(GTValueArray [i]).AsPLs);
}
else if (gtKey.Equals(VCFConstants.DEPTH_KEY))
{
gb.DP = (Convert.ToInt32(GTValueArray [i]));
}
else
{
gb.AddAttribute(gtKey, GTValueArray [i]);
}
}
}
}
List <Allele> GTalleles;
if (genotypeAlleleLocation == -1)
{
GTalleles = new List <Allele> (0);
}
else
{
GTalleles = parseGenotypeAlleles(GTValueArray [genotypeAlleleLocation], alleles, alleleMap);
}
gb.Alleles = GTalleles;
gb.Phased = genotypeAlleleLocation != -1 && GTValueArray [genotypeAlleleLocation].IndexOf(VCFConstants.PHASED_AS_CHAR) != -1;
// add it to the list
try {
curGenotype = gb.Make();
} catch (Exception e) {
throw new VCFParsingError(e.Message + ", at position " + chr + ":" + pos);
}
}
genotypes.Add(curGenotype);
}
return(new LazyGenotypesContext.LazyData(genotypes, header.SampleNamesInOrder, header.SampleNameToOffset));
}
/// <summary>
/// Create a genotype map
/// </summary>
/// <param name="str"> the string </param>
/// <param name="alleles"> the list of alleles </param>
/// <returns> a mapping of sample name to genotype object </returns>
public LazyGenotypesContext.LazyData CreateGenotypeMap (string str, IList<Allele> alleles, string chr, int pos)
{
if (genotypeParts == null)
genotypeParts = new String[header.ColumnCount - NUM_STANDARD_FIELDS];
try {
FastStringUtils.Split (str, VCFConstants.FIELD_SEPARATOR_CHAR, genotypeParts);
} catch (Exception e) {
throw new VCFParsingError ("Could not parse genotypes, was expecting " + (genotypeParts.Length - 1).ToString () + " but found " + str.Split (VCFConstants.FIELD_SEPARATOR_CHAR).Length.ToString (), e);
}
List<Genotype> genotypes = new List<Genotype> (genotypeParts.Length);
// get the format keys
//int nGTKeys = ParsingUtils.Split(genotypeParts[0], genotypeKeyArray, VCFConstants.GENOTYPE_FIELD_SEPARATOR_CHAR);
string[] genotypeKeyArray = genotypeParts [0].Split (VCFConstants.GENOTYPE_FIELD_SEPARATOR_CHAR);
int genotypeAlleleLocation = Array.IndexOf (genotypeKeyArray, VCFConstants.GENOTYPE_KEY);
if (version != VCFHeaderVersion.VCF4_1 && genotypeAlleleLocation == -1) {
generateException ("Unable to find the GT field for the record; the GT field is required in VCF4.0");
}
// clear out our allele mapping
alleleMap.Clear ();
GenotypeBuilder gb = new GenotypeBuilder ();
// cycle through the genotype strings
for (int genotypeOffset = 1; genotypeOffset < genotypeParts.Length; genotypeOffset++) {
Genotype curGenotype;
string sampleName = header.GenotypeSampleNames [genotypeOffset - 1];
var currentGeno = genotypeParts [genotypeOffset];
//shortcut for null alleles
if (currentGeno == "./.") {
curGenotype = GenotypeBuilder.CreateMissing (sampleName, 2);
} else if (currentGeno == ".") {
curGenotype = GenotypeBuilder.CreateMissing (sampleName, 1);
} else {
gb.Reset (false);
gb.SampleName = sampleName;
string[] GTValueArray = FastStringUtils.Split (currentGeno, VCFConstants.GENOTYPE_FIELD_SEPARATOR_CHAR, int.MaxValue, StringSplitOptions.None);
// cycle through the sample names
// check to see if the value list is longer than the key list, which is a problem
if (genotypeKeyArray.Length < GTValueArray.Length) {
generateException ("There are too many keys for the sample " + sampleName + ", line is: keys = " + genotypeParts [0] + ", values = " + genotypeParts [genotypeOffset]);
}
if (genotypeAlleleLocation > 0) {
generateException ("Saw GT field at position " + genotypeAlleleLocation + ", but it must be at the first position for genotypes when present");
}
//TODO: THIS IS A DAMNED MESS
//Code loops over all fields in the key and decodes them, adding them as information to the genotype builder, which then makes it.
if (genotypeKeyArray.Length > 0) {
gb.MaxAttributes (genotypeKeyArray.Length - 1);
for (int i = 0; i < genotypeKeyArray.Length; i++) {
string gtKey = genotypeKeyArray [i];
if (i >= GTValueArray.Length) {
break;
}
// todo -- all of these on the fly parsing of the missing value should be static constants
if (gtKey == VCFConstants.GENOTYPE_FILTER_KEY) {
IList<string> filters = parseFilters (GetCachedString (GTValueArray [i]));
if (filters != null) {
gb.SetFilters (filters.ToList ());
}
} else if (GTValueArray [i] == VCFConstants.MISSING_VALUE_v4) {
// don't add missing values to the map
} else {
if (gtKey == VCFConstants.GENOTYPE_QUALITY_KEY) {
if (GTValueArray [i] == VCFConstants.MISSING_GENOTYPE_QUALITY_v3) {
gb.noGQ ();
} else {
gb.GQ = ((int)Math.Round (Convert.ToDouble (GTValueArray [i])));
}
} else if (gtKey == VCFConstants.GENOTYPE_ALLELE_DEPTHS) {
gb.AD = (decodeInts (GTValueArray [i]));
} else if (gtKey == VCFConstants.GENOTYPE_PL_KEY) {
gb.PL = (decodeInts (GTValueArray [i]));
} else if (gtKey == VCFConstants.GENOTYPE_LIKELIHOODS_KEY) {
gb.PL = (GenotypeLikelihoods.fromGLField (GTValueArray [i]).AsPLs);
} else if (gtKey.Equals (VCFConstants.DEPTH_KEY)) {
gb.DP = (Convert.ToInt32 (GTValueArray [i]));
} else {
gb.AddAttribute (gtKey, GTValueArray [i]);
}
}
}
}
List<Allele> GTalleles;
if (genotypeAlleleLocation == -1) {
GTalleles = new List<Allele> (0);
} else {
GTalleles = parseGenotypeAlleles (GTValueArray [genotypeAlleleLocation], alleles, alleleMap);
}
gb.Alleles = GTalleles;
gb.Phased = genotypeAlleleLocation != -1 && GTValueArray [genotypeAlleleLocation].IndexOf (VCFConstants.PHASED_AS_CHAR) != -1;
// add it to the list
try {
curGenotype = gb.Make ();
} catch (Exception e) {
throw new VCFParsingError (e.Message + ", at position " + chr + ":" + pos);
}
}
genotypes.Add (curGenotype);
}
return new LazyGenotypesContext.LazyData (genotypes, header.SampleNamesInOrder, header.SampleNameToOffset);
}
