/**
* Do the full tangent normalization process given proportional coverage data.
*
* <p>This includes:
*
* <ul>
* <li>normalization by target factors
* <li>projection of the normalized coverage profile into the hyperplane from the PoN
* </ul>
*
* @param pon -- never {@code null}
* @param pcov -- never {@code null}. Must contain data for at least one sample.
* @param ctx spark context. Use {@code null} if no context is available
* @return never {@code null}
*/
public static TangentNormalizationResult tangentNormalizePcov(
final PoN pon, final ReadCountCollection pcov, final JavaSparkContext ctx) {
Utils.nonNull(pon, "PoN cannot be null.");
Utils.nonNull(pcov, "input pcov read counts cannot be null when creating a coverage profile.");
ParamUtils.isPositive(
pcov.columnNames().size(), "input cov profile column names cannot be an empty list.");
final ReadCountCollection coverageProfile = createCoverageProfile(pon, pcov);
return TangentNormalizer.tangentNormalize(pon, coverageProfile, ctx);
}
/**
* Project all of the normals used to create the PoN into the reduced panel.
*
* @param pon Not {@code null}.
* @param ctx Spark context, {@code null} indicates no spark context available, which is
* supported.
* @return never {@code null}. Result will contain multiple columns in each of the
* ReadCountCollection attributes.
*/
public static TangentNormalizationResult tangentNormalizeNormalsInPoN(
final PoN pon, final JavaSparkContext ctx) {
// Get the list of sample names in a modifiable List
final List<String> sampleNamesCopy = new ArrayList<>(pon.getSampleNames());
logger.info("Loading normalized counts...");
final ReadCountCollection coverageProfile =
new ReadCountCollection(
SetUniqueList.setUniqueList(pon.getTargets()),
SetUniqueList.setUniqueList(sampleNamesCopy),
pon.getNormalizedCounts());
logger.info("Tangent normalizing the normals (normalized by target factors) ...");
// For each sample in the PoN, tangent normalize against the qc reduced PoN.
return TangentNormalizer.tangentNormalize(pon, coverageProfile, ctx);
}
/**
* {@link TangentNormalizer :: tangentNormalizePcov} with no spark context
*
* @param pon never {@code null}. Must contain data for at least one sample.
* @param pcov never {@code null}. Must contain data for at least one sample.
* @return never {@code null}
*/
public static TangentNormalizationResult tangentNormalizePcov(
final PoN pon, final ReadCountCollection pcov) {
return TangentNormalizer.tangentNormalizePcov(pon, pcov, null);
}