Python tutorials

Exporting data

Learn how to stream the single-cell data and metadata from Census into your machine.

• Quick start
• Querying and fetching the single-cell data and cell/gene metadata.
• Generating citations for Census slices
• Querying data using the gget cellxgene module

[NEW! 🚀] Using integrated embeddings and models

Tutorials that show you how to retrieve pre-calculated Census embeddings and use their associated models for your workflows.

Access Census embeddings.

• Access CELLxGENE collaboration embeddings (scVI, Geneformer)
• Access CELLxGENE-hosted embeddings
• Find the most similar Census cells in any other data with embeddings vector search

Use the Census trained models.

• scVI for cell type prediction and data projection
• Geneformer for cell class prediction and data projection

Exploring human biology with Census embeddings.

• Exploring biologically relevant clusters in Census embeddings

Understanding Census data

Gain a better understanding on the nature of the Census data and how it’s organized.

• Understanding and filtering out duplicate cells
• Learning about the CZ CELLxGENE Census
• Exploring all data from a tissue
• Exploring the Census Datasets table
• Genes measured in each cell (dataset presence matrix)
• Exploring pre-calculated summary cell counts

Analyzing Census data

A few examples of relevant analysis pipelines with Census data.

• Summarizing cell and gene metadata
• Integrating multi-dataset slices of data
• Normalizing full-length gene sequencing data

Scalable computing

Demonstrations of memory-efficient compute workflows that leverage the streaming capabilities of Census.

• Experimental Highly Variable Genes API
• Out-of-core (incremental) mean and variance calculation
• Computing on X using online (incremental) algorithms

Scalable machine learning

Learn about features to do data modeling directly from Census into machine learning toolkits.

• Training a PyTorch Model