Quick Start#

Install & import the library#

pip install adata-query

import adata_query

AnnData#

This package is downstream of data loading and assumes a generally typical implementation of adata created using the AnnData package or Scanpy.

import anndata

h5ad_path = "/path/to/your/adata.h5ad"

adata = anndata.read_h5ad(h5ad_path)

Once you have some data, you are ready to interface with adata_query.

adata_query.fetch#

The fetch function is probably the most useful function in the library. It’s also the most function most likely to be touched by the user. It relies on the other two functions in the library: locate and format_data. (described briefly, below).

In short, this function finds a matrix stored in adata using str keyword. Importantly, this function allows the user to do this in a grouped fashion, based on pd.groupby

Ungrouped
key = "X_pca" # stored in adata.obsm

data = adata_query.fetch(adata = adata, key = "X_pca")
Grouped
key = "X_pca" # stored in adata.obsm
groupby = "cluster" # cell annotation in adata.obs

data = adata_query.fetch(
   adata = adata,
   key = key,
   groupby = groupby,
)

In this example, data is returned as List.

adata_query.format_data#

This function allows us to automatically format data stored as np.ndarray as a torch.Tensor, on any device.

numpy → numpy
data = adata_query.format(data) # returns np.ndarray
numpy → torch (cpu)
data = adata_query.format(data, torch = True, device = "cpu") # torch.Tensor on cpu
numpy → torch (gpu)
data = adata_query.format(data, torch = True) # torch.Tensor on gpu, if available

# torch.Tensor can also be explicitly declared to a specific device
data = adata_query.format(data, torch = True, device = "cuda:0")

# Apple Silicon also works and will be automatically detected
data = adata_query.format(data, torch = True, device = "mps:0")

adata_query.locate#

This function simply returns the sub-container location of a matrix, given its accessor key. While useful in the implementation of the adata_query.fetch function, it is not anticipated to be widely-used beyond that scope.

.. code-block:: python

import adata_query

key = “X_pca” attr_key = adata_query.locate(adata, key = key) # attr_key = “obsm”

Note

While both the format_data and locate functions may seem trivial, they are useful in adding flexibility to complex workflows.