Welcome to scAtlasVAE’s documentation!

Contents

• Introduction
  • Installation
    • PyTorch installation
  • Tutorial
    • Basic Usage
  • Key functionality
  • Model architecture
• Preprocessing Immune Profiling Data
• Automatic transfer to pan-disease CD8+ T cell atlas
  • Step1: Load the reference and query data
  • Step2: Transfer the cell type annotation
  • Step3: Transfer the higher-resolution cell type annotation for Tex
• Integrating Multi-source Gene Expression (GEX)
  • Training the VAE model using batch key
  • Training the VAE model using batch key and categorical covariates (e.g. study_name)
  • Training the VAE model using batch key and label key (e.g. cell_type)
  • Training the VAE model using multiple batch keys and mutiple label keys
  • Saving the VAE model
• Transfering Multi-source Gene Expression (GEX)
  • Transfer without training query datasets with reference datasets
    • Getting the transfered latent embedding
    • Mapping the UMAP representation to the reference
  • Transfer by training query datasets with reference datasets
• Re-Training Gene Expression (GEX)
• API
  • Preprocessing
    • VDJPreprocessingV1Human
      • VDJPreprocessingV1Human.process()
    • VDJPreprocessingV1Mouse
      • VDJPreprocessingV1Mouse.process()
  • Model
    • scAtlasVAE
    • scAtlasVAE’s Methods table
    • scAtlasVAE’s Methods
      • scAtlasVAE.partial_load_state_dict()
      • scAtlasVAE.get_config()
      • scAtlasVAE.save_to_disk()
      • scAtlasVAE.load_from_disk()
      • scAtlasVAE.setup_anndata()
      • scAtlasVAE.fit()
  • Pipeline
    • pipeline.run_transfer()
  • Tools
    • umap_alignment()
    • cell_type_alignment()
  • Utilities
    • get_default_device()