crested.tl.extract_layer_embeddings

crested.tl.extract_layer_embeddings#

crested.tl.extract_layer_embeddings(input, model, layer_name, genome=None, **kwargs)#

Extract embeddings from a specified layer for all inputs.

Parameters:

  • input (str | list[str] | ndarray | AnnData) – Input data to get embeddings for. Can be a (list of) sequence(s), a (list of) region name(s), a matrix of one hot encodings (N, L, 4), or an AnnData object with region names as its var_names.

  • model (Model) – A trained keras model from which to extract the embeddings.

  • layer_name (str) – The name of the layer from which to extract the embeddings.

  • genome (Union[Genome, PathLike, None] (default: None)) – Genome or path to the genome fasta. Required if no genome is registered and input is an anndata object or region names.

  • **kwargs – Additional keyword arguments to pass to the keras.Model.predict method.

Return type:

ndarray

Returns:

Embeddings of shape (N, D), where N is the number of regions in the input and D is the size of the embedding layer.

Contents