Package org.broadinstitute.hellbender.tools.walkers.vqsr.scalable.modeling
package org.broadinstitute.hellbender.tools.walkers.vqsr.scalable.modeling
-
ClassDescriptionGiven an HDF5 file containing annotations for a training set (in the format specified by
VariantAnnotationsModel.trainAndSerialize(java.io.File, java.lang.String)), a Python script containing modeling code, and a JSON file containing hyperparameters, thePythonSklearnVariantAnnotationsModel.trainAndSerialize(java.io.File, java.lang.String)method can be used to train a model.Given an HDF5 file containing annotations for a test set (in the format specified byVariantAnnotationsScorer.score(java.io.File, java.io.File)), a Python script containing scoring code, and a file containing a pickled Python lambda function for scoring, thePythonSklearnVariantAnnotationsScorer.score(java.io.File, java.io.File)method can be used to generate scores.File interface for passing annotations to a modeling backend and indicating a path prefix for resulting output.File interface for passing annotations to a scoring backend and returning scores.