scvelo.tl.differential_kinetic_test

scvelo.tl.differential_kinetic_test(data, var_names='velocity_genes', groupby=None, use_raw=None, return_model=None, add_key='fit', copy=None, **kwargs)

Test to detect cell types / lineages with different kinetics.

Likelihood ratio test for differential kinetics to detect clusters/lineages that display kinetic behavior that cannot be sufficiently explained by a single model for the overall dynamics. Each cell type is tested whether an independent fit yields a significantly improved likelihood.

https://user-images.githubusercontent.com/31883718/78930730-dc737200-7aa4-11ea-92f6-269b7609c3a5.png
Parameters:
  • data (AnnData) – Annotated data matrix.

  • var_names (str, list of str (default: ‘velocity_genes)) – Names of variables/genes to use for the fitting.

  • groupby (str (default: None)) – Key of observations grouping to consider, e.g. ‘clusters’.

  • use_raw (bool (default: False)) – Whether to use raw data for estimation.

  • add_key (str (default: ‘fit’)) – Key to add to parameter names, e.g. ‘fit_t’ for fitted time.

  • copy (bool (default: None)) – Return a copy instead of writing to adata.

Returns:

  • fit_pvals_kinetics (.varm) – P-values of competing kinetic for each group and gene

  • fit_diff_kinetics (.var) – Groups that have differential kinetics for each gene.