public void testRemoveMulti() throws Exception {
try (TestDirectory td = new TestDirectory()) {
final File f =
FileHelper.resourceToGzFile(
"com/rtg/vcf/resources/vcfsubset.vcf", new File(td, "vcf.vcf.gz"));
final File out = new File(td, "out.vcf");
checkMainInitOk(
"-i",
f.getPath(),
"-o",
out.getPath(),
"--remove-samples",
"--keep-info",
"AN",
"--keep-info",
"AC",
"--keep-filter",
"YEA",
"-Z");
final String content = FileHelper.fileToString(out);
final String editedContent =
StringUtils.grepMinusV(StringUtils.grepMinusV(content, "^##RUN-ID"), "^##CL");
mNano.check("vcfsubset-multi.vcf", editedContent);
}
}
public void testExplosion() throws Exception {
try (TestDirectory td = new TestDirectory()) {
final File f =
FileHelper.resourceToGzFile(
"com/rtg/vcf/resources/vcfsubset.vcf", new File(td, "vcf.vcf.gz"));
final File out = new File(td, "out.vcf");
assertEquals(
"Records skipped due to invalid or incompatible sample fields: 1" + StringUtils.LS,
checkMainInitWarn(
"-i",
f.getPath(),
"-o",
out.getPath(),
"--remove-format",
"GT",
"--remove-format",
"DS",
"-Z"));
final String content = FileHelper.fileToString(out);
final String nonheader = StringUtils.grepMinusV(content, "^#");
assertTrue(nonheader.startsWith("X\t60052"));
}
}
private static String createVariantName(
final int leftClip, final int rightClip, final String oldName) {
if (oldName == null) {
return oldName;
}
final int colon = oldName.indexOf(VariantUtils.COLON);
if (colon >= 0) {
assert colon >= 0 && oldName.indexOf(VariantUtils.COLON, colon + 1) == -1;
final String allele1 = StringUtils.clip(oldName.substring(0, colon), leftClip, rightClip);
final String allele2 = StringUtils.clip(oldName.substring(colon + 1), leftClip, rightClip);
return allele1 + VariantUtils.COLON + allele2;
} else {
assert oldName.indexOf(VariantUtils.COLON) == -1;
return StringUtils.clip(oldName, leftClip, rightClip);
}
}
private static VariantLocus createLocus(
final Variant original, final int leftClip, final int rightClip) {
final VariantLocus locus = original.getLocus();
final char newPrevNt =
leftClip == 0 ? locus.getPreviousRefNt() : locus.getRefNts().charAt(leftClip - 1);
final String newReference = StringUtils.clip(locus.getRefNts(), leftClip, rightClip);
final int newStart = locus.getStart() + leftClip;
final int newEnd = locus.getEnd() - rightClip;
return new VariantLocus(locus.getSequenceName(), newStart, newEnd, newReference, newPrevNt);
}
// Can't use Description, as that includes hypotheses that weren't necessarily called.
private static HashSet<String> extractCalledAlleles(final Variant variant) {
final HashSet<String> alleles = new HashSet<>();
for (int k = 0; k < variant.getNumberOfSamples(); k++) {
final VariantSample vs = variant.getSample(k);
if (vs != null) {
if (!vs.isIdentity()) {
Collections.addAll(alleles, StringUtils.split(vs.getName(), VariantUtils.COLON));
}
}
}
return alleles;
}
void infoString(final StringBuilder sb) {
sb.append("Memory Usage: ").append(mNumberSequences).append(" sequences").append(LS);
long totalBytes = mData.infoString(sb);
if (mNames != null) {
sb.append("\t\t")
.append(StringUtils.commas(mNames.bytes()))
.append("\t")
.append(StringUtils.commas(mNames.length()))
.append("\tNames")
.append(LS);
totalBytes += mNames.bytes();
}
if (mNameSuffixes != null) {
sb.append("\t\t")
.append(StringUtils.commas(mNameSuffixes.bytes()))
.append("\t")
.append(StringUtils.commas(mNameSuffixes.length()))
.append("\tSuffixes")
.append(LS);
totalBytes += mNameSuffixes.bytes();
}
sb.append("\t\t").append(StringUtils.commas(totalBytes)).append("\t\tTotal bytes").append(LS);
}
public void testKeepSamples() throws Exception {
try (TestDirectory td = new TestDirectory()) {
final File f =
FileHelper.resourceToGzFile(
"com/rtg/vcf/resources/vcfsubset.vcf", new File(td, "vcf.vcf.gz"));
final File out = new File(td, "out.vcf");
checkMainInitOk(
"-i",
f.getPath(),
"-o",
out.getPath(),
"--keep-sample",
"HG00096",
"--keep-sample",
"HG00100",
"-Z");
final String content = FileHelper.fileToString(out);
final String editedContent =
StringUtils.grepMinusV(StringUtils.grepMinusV(content, "^##RUN-ID"), "^##CL");
mNano.check("vcfsubset-keepsamples.vcf", editedContent);
}
}
@Override
public void toString(final StringBuilder sb) {
sb.append("Contaminated Cancer Model");
final FormatReal fmt = new FormatReal(4, 3);
sb.append(" contamination=").append(fmt.format(mContamination));
sb.append(LS);
final int pad = hypotheses().nameLength();
final int size = ((HypothesesCancer) hypotheses()).subHypotheses().size();
for (int i = 0; i < size; i++) {
sb.append(StringUtils.padLeft(hypotheses().name(i), pad));
for (int j = 0; j < size; j++) {
final int k = hypotheses().code().code(i, j);
sb.append(fmt.format(arithmetic().poss2Ln(mPosteriors[k])));
}
sb.append(LS);
}
}
private static Map<Set<String>, Double> newGenotypeLikelihoods(
int leftClip, int rightClip, VariantSample sample) {
final Map<Set<String>, Double> newMap = new HashMap<>();
final Map<Set<String>, Double> originalLikelihoods = sample.getGenotypeLikelihoods();
if (originalLikelihoods != null) {
for (Map.Entry<Set<String>, Double> entry : originalLikelihoods.entrySet()) {
final Set<String> newSet = new HashSet<>();
for (String s : entry.getKey()) {
newSet.add(StringUtils.clip(s, leftClip, rightClip));
}
final Double v = newMap.get(newSet);
final double existing = v == null ? LogApproximatePossibility.SINGLETON.zero() : v;
newMap.put(newSet, LogApproximatePossibility.SINGLETON.add(existing, entry.getValue()));
}
return newMap;
} else {
return null;
}
}
private static Variant createSplitVariant(
final Variant original, final int start, final int end, final int id) {
final VariantLocus newLocus = createLocus(original, start, end);
final VariantSample[] newSamples;
if (original.getNumberOfSamples() > 0) {
// trim description
Description oldDescription = DescriptionNone.SINGLETON;
for (int i = 0; i < original.getNumberOfSamples(); i++) {
if (original.getSample(i) != null
&& !(original.getSample(i).getStats().counts().getDescription()
instanceof DescriptionNone)) {
oldDescription = original.getSample(i).getStats().counts().getDescription();
}
}
// Incrementally build up mapping of old alleles to new alleles
final LinkedHashMap<String, Integer> alleles = new LinkedHashMap<>();
final int[] alleleMap = new int[oldDescription.size()];
for (int i = 0; i < oldDescription.size(); i++) {
final String clipped = StringUtils.clip(oldDescription.name(i), start, end);
Integer newPos = alleles.get(clipped);
if (newPos == null) {
newPos = alleles.size();
}
alleles.put(clipped, newPos);
alleleMap[i] = newPos;
}
final Description newDescription =
new DescriptionCommon(alleles.keySet().toArray(new String[alleles.size()]));
newSamples = createVariants(original, start, end, newDescription, alleleMap);
} else {
newSamples = new VariantSample[0];
}
final Variant result = new Variant(newLocus, newSamples);
Variant.copy(original, result);
result.setPossibleCause(createVariantName(start, end, original.getPossibleCause()));
result.setSplitId(id);
return result;
}
/**
* Trim a variant by removing common prefix and suffix from each call and reference and adjusting
* all the position information accordingly.
*
* @param original the original call
* @return trimmed variant (possibly original if no change made)
*/
static Variant trim(final Variant original) {
final String ref = original.getLocus().getRefNts();
if (ref.length() == 0) {
// Cannot possibly trim if we are inserting, all bases are to be inserted
return original;
}
// Compute set of called alleles and exit if there are none
final HashSet<String> catSet = extractCalledAlleles(original);
if (catSet.size() == 0) {
return original;
}
// Include actual reference sequence (this will not be "")
assert ref.length() > 0;
catSet.add(ref);
// Compute maximal clip positions based on set of alleles
final String[] cats = catSet.toArray(new String[catSet.size()]);
final int rightClip = StringUtils.longestSuffix(cats);
final int leftClip = StringUtils.longestPrefix(rightClip, cats);
// final int leftClip = StringUtils.longestPrefix(cats);
// final int rightClip = StringUtils.longestSuffix(cats, leftClip);
// Quick exit if no trimming is possible
if (leftClip == 0 && rightClip == 0) {
return original;
}
// Create new locus, will be "" for case of pure insertion
final VariantLocus newLocus = createLocus(original, leftClip, rightClip);
final VariantSample[] newSamples;
if (original.getNumberOfSamples() > 0) {
// trim description
Description oldDescription = DescriptionNone.SINGLETON;
for (int i = 0; i < original.getNumberOfSamples(); i++) {
if (original.getSample(i) != null
&& !(original.getSample(i).getStats().counts().getDescription()
instanceof DescriptionNone)) {
oldDescription = original.getSample(i).getStats().counts().getDescription();
}
}
final String[] alleles = new String[oldDescription.size()];
final int[] alleleMap = new int[oldDescription.size()];
for (int i = 0; i < oldDescription.size(); i++) {
final String clipped = StringUtils.clip(oldDescription.name(i), leftClip, rightClip);
alleles[i] = clipped;
alleleMap[i] = i;
}
final Description newDescription = new DescriptionCommon(alleles);
// Create new trimmed samples
newSamples = createVariants(original, leftClip, rightClip, newDescription, alleleMap);
} else {
newSamples = new VariantSample[0];
}
// Create new variant from the new samples
final Variant result = new Variant(newLocus, newSamples);
Variant.copy(original, result);
// This must be done after the copy
result.setPossibleCause(createVariantName(leftClip, rightClip, original.getPossibleCause()));
result.setTrimmed();
// System.err.println(original);
// System.err.println(result);
return result;
}