cellrank.datasets.pancreas(path='datasets/endocrinogenesis_day15.5.h5ad', kind='raw', **kwargs)[source]

Development of the murine pancreas at E15.5 from [Bastidas-Ponce et al., 2019].

scRNA-seq dataset containing 2531 cells recorded using 10x Chromium in a single time point. Data was filtered to remove heavily cycling populations and to focus on the late stages of endocrinogenesis.

  • path (Union[str, Path]) – Path where to save the dataset.

  • kind (Literal['raw', 'preprocessed', 'preprocessed-kernel']) –

    What kind of dataset to download. Valid option are:

    • 'raw' - contains raw spliced and unspliced count data, low-dimensional embedding coordinates as well as original cluster annotations.

    • 'preprocessed' - same as above, but additionally genes have been filtered to contain at least 20 unspliced and spliced counts and subsetted to the top 2000 highly variable genes. Data has been normalized by total counts and log-transformed.

    • 'preprocessed-kernel' - same as above, but additionally a cell-cell transition matrix has been computed using the VelocityKernel.

  • kwargs (Any) – Keyword arguments for read().

Return type:



: Annotated data object.