cpa.ComPertAPI.predict

ComPertAPI.predict(genes, df, uncertainty=True, sample=False, n_samples=10)

Predict values of control ‘genes’ conditions specified in df.

Parameters:

genes np.array

Control cells.

df pd.DataFrame

Values for perturbations and covariates to generate.

uncertainty bool (default: True)

Compute uncertainties for the generated cells.

sample bool (default: False)

If sample is True, returns samples from gausssian distribution with mean and variance estimated by the model. Otherwise, returns just means and variances estimated by the model.

n_samples int (default: 10)

Number of samples to sample if sampling is True.

Returns:

: If return_anndata is True, returns anndata structure. Otherwise, returns np.arrays for gene_means, gene_vars and a data frame for the corresponding conditions df_obs.