taxus 0.0.4

Gaussian Process models for transcriptome data

taxus

Gaussian Process models for transcriptome data

pip install taxus

import taxus as tx

# at the moment importing data has to be handled by the user
covariates, counts = get_mock_data()
gp = tx.GP('~ time + treatment', covariates, counts, kernel='rbf', likelihood='poisson')
elbo = gp.fit()

likelihood_ratio_rbf = tx.LRT(
    full_formula='~ time + treatment',
    reduced_formula='~ time',
    covariates=covariates,
    expression=counts,
    kernel='rbf',
    likelihood='nb'
)

likelihood_ratio_linear = tx.LRT(
    full_formula='~ C(time) + C(treatment) + C(time) : C(treatment)',
    reduced_formula='~ C(time) + C(treatment)',
    covariates=covariates,
    expression=counts,
    kernel='linear',
    likelihood='nb'
)