omicverse.single.cytotrace2

omicverse.single.cytotrace2(adata, use_model_dir, species='mouse', batch_size=10000, smooth_batch_size=1000, disable_parallelization=False, max_cores=None, max_pcs=200, seed=14, output_dir='cytotrace2_results')[source]

Predict developmental potency with CytoTRACE2.

Parameters:

  • adata (anndata.AnnData) – Single-cell expression object. Uses adata.X as the expression matrix.

  • use_model_dir (str) – Directory containing CytoTRACE2 pretrained model weights.

  • species (str, default='mouse') – Species label used by gene preprocessing. Supported values are typically 'mouse' and 'human'.

  • batch_size (int, default=10000) – Number of cells per prediction batch.

  • smooth_batch_size (int, default=1000) – Number of cells per smoothing batch in diffusion/KNN post-processing.

  • disable_parallelization (bool, default=False) – Whether to disable multiprocessing.

  • max_cores (int or None, default=None) – Maximum cores used when parallelization is enabled. None lets the function infer an available value.

  • max_pcs (int, default=200) – Number of principal components used in KNN smoothing.

  • seed (int, default=14) – Random seed used for chunking and stochastic operations.

  • output_dir (str, default='cytotrace2_results') – Directory for intermediate files and final result table.

Returns:

CytoTRACE2 result table indexed by cell ID, including CytoTRACE2_Score, potency category, and relative rank score.

Return type:

pandas.DataFrame

Notes

This function also writes prediction columns into adata.obs.