Unifying base data object

[RemyLau]

Currently, there are several different dataset objects specialized for each task and model (e.g., CellTypeDataset, ClusteringDataset), each of them takes a variety of specialized arguments that are not directly related to the underlying data, e.g., save path, processing scheme, choice of tissue. This complexity makes it quite hard to maintain the code base and implement new methods/datasets.

To improve this situation, we need to isolate raw dataset objects from transformation/processing methods.

• Base data object
  • Take AnnData as an input and save it as a private attribute (read-only?).
  • Construct data loaders that load g, x, y, etc., to be passed to the model for training/evaluation.
• Dataset object
  • Download option
  • Transformation option
  • Dataset from paper (preprocessed) -> used to benchmark the reproducibility of the reimplemented model
• Transformation
  • Leverage functionalities from scanpy (recall that now the base data object store an AnnData object as a (private) attribute

To fix

Single modality

• examples/single_modality/clustering/scdsc.py (Update clustering examples to use data object #95)
• examples/single_modality/clustering/graphsc.py (Update clustering examples to use data object #95)
• examples/single_modality/clustering/scdcc.py (Update clustering examples to use data object #95)
• examples/single_modality/clustering/sctag.py (Update clustering examples to use data object #95)
• examples/single_modality/clustering/scdeepcluster.py (Update clustering examples to use data object #95)
• examples/single_modality/imputation/graphsci.py
• examples/single_modality/imputation/deepimpute.py
• examples/single_modality/imputation/scgnn.py
• examples/single_modality/cell_type_annotation/singlecellnet.py (update singlecellnet to use dance data object #77)
• examples/single_modality/cell_type_annotation/celltypist.py (update celltypist example script to use dance data object #72)
• examples/single_modality/cell_type_annotation/scdeepsort.py (update scDeepSort example script to use dance data object #75)
• examples/single_modality/cell_type_annotation/actinn.py (update actinn example script with new data object #63)
• examples/single_modality/cell_type_annotation/svm.py (refactor data object and update svm #56, refactor label adata processing to increase compatibility with user defined datasets #57)

Spatial

• examples/spatial/spatial_domain/stagate.py (update stagate example script to use dance data object #127)
• examples/spatial/spatial_domain/louvain.py (update louvain example script to use dance dataobject #124)
• examples/spatial/spatial_domain/stlearn.py (update stlearn example script to use dance data object  #126)
• examples/spatial/spatial_domain/spagcn.py (refactor spagcn graph construct; use dance data object in the example spagcn script #83)
• examples/spatial/cell_type_deconvo/spotlight.py (update spotlight example script to use dance data object #107)
• examples/spatial/cell_type_deconvo/dstg.py (update dstg example script to use dance data, refactor dstg graph construct #103)
• examples/spatial/cell_type_deconvo/card.py (update card cell-type deconvolution example script using dance data object #93)
• examples/spatial/cell_type_deconvo/spatialdecon.py (update spatialdeconv to use dance data object #94)

Multi modality

• examples/multi_modality/joint_embedding/scmvae.py
• examples/multi_modality/joint_embedding/dcca.py
• examples/multi_modality/joint_embedding/jae.py
• examples/multi_modality/joint_embedding/scmogcnv2.py
• examples/multi_modality/joint_embedding/scmogcn.py
• examples/multi_modality/match_modality/cmae.py
• examples/multi_modality/match_modality/scmm.py
• examples/multi_modality/match_modality/scmogcn.py
• examples/multi_modality/predict_modality/babel.py (update babel example script to use dance data object #89)
• examples/multi_modality/predict_modality/cmae.py
• examples/multi_modality/predict_modality/scmm.py
• examples/multi_modality/predict_modality/scmogcn.py