v2.0.0 renaming and deprecation

With CREsted v2.0.0, we’ve reorganised the plotting functions, to align by what they plot rather than how they plot. All functions are still there, often with expanded functionality, and new ones have been added.

Old name	New name
bar.region	region.bar
bar.region_predictions	region.bar
bar.normalization_weights	qc.normalization_weights
hist.distribution	dist.histogram
heatmap.correlations_self	corr.heatmap_self
heatmap.correlations_predictions	corr.heatmap
locus.locus_scoring	unchanged
locus.track	unchanged
scatter.class_density	corr.scatter
violin.correlations	corr.violin
patterns.contribution_scores	explain.contribution_scores
patterns modisco functions	modisco.*
patterns.enhancer_design_steps_contribution_scores	design.step_contribution_scores
patterns.enhancer_design_steps_predictions	design.step_predictions

We’ve also slightly reorganised the enhancer design functions in tl:

Old name	New name
tl.enhancer_design_in_silico_evolution	tl.design.in_silico_evolution
tl.enhancer_design_motif_insertion	tl.design.motif_insertion
utils.EnhancerOptimizer	tl.design.EnhancerOptimizer
utils.derive_intermediate_sequences	tl.design.derive_intermediate_sequences

We’ve also cleaned up functions and methods that have long been marked as deprecated:

• The methods of Crested which were superseded by tl functions:

  • get_embeddings -> extract_layer_embeddings()

  • predict -> predict()

  • predict_regions -> predict()

  • predict_sequence -> predict()

  • score_gene_locus -> score_gene_locus()

  • calculate_contribution_scores -> contribution_scores()

  • calculate_contribution_scores_regions -> contribution_scores()

  • calculate_contribution_scores_sequence -> contribution_scores()

  • calculate_contribution_scores_enhancer_design -> contribution_scores()

  • tfmodisco_calculate_and_save_contribution_scores_sequences -> contribution_scores_specific()

  • tfmodisco_calculate_and_save_contribution_scores -> contribution_scores_specific()

  • enhancer_design_motif_implementation -> motif_insertion()

  • enhancer_design_in_silico_evolution -> in_silico_evolution()

  • _derive_intermediate_sequences -> derive_intermediate_sequences()

• Aliases for models that didn’t properly reflect their nature:

  • chrombpnet -> dilated_cnn()

  • chrombpnet_decoupled -> dilated_cnn_decoupled()

• Superseded or obsolete utility functions:

  • extract_bigwig_values_per_bp -> read_bigwig_region()

  • get_value_from_dataframe -> df.loc[row_name, column_name]