Tools: tl

tl.PILOT(sample_key, cell_group_key, ...[, ...])	Optimal transport based method to compute the Wasserstein distance between two single single-cell experiments.
tl.GroupedPseudobulk(sample_key, cell_group_key)	Baseline, where distances between samples are average distances between their cell group pseudobulks
tl.CellGroupComposition(sample_key, ...[, ...])	A simple baseline, which represents samples as composition of their cell groups (for example, cell type fractions)
tl.MrVI(sample_key, cell_group_key[, ...])	Deep generative modeling for quantifying sample-level heterogeneity in single-cell omics.
tl.RandomVector(sample_key, cell_group_key)	A dummy baseline, which represents samples as random embeddings
tl.SCPoli(sample_key, cell_group_key[, ...])	A semi-supervised conditional deep generative model from https://www.biorxiv.org/content/10.1101/2022.11.28.517803v1
tl.Pseudobulk(sample_key, cell_group_key[, ...])	A simple baseline, which represents samples as pseudobulk of their gene expression
tl.WassersteinTSNE(sample_key, ...[, layer, ...])	Method based on the matrix of pairwise Wasserstein distances between units.
tl.SampleRepresentationMethod(sample_key, ...)	Base class for sample representation methods
tl.describe_metadata(metadata)	Prints the basic information about the metadata and tries to guess column types
tl.test_distances_significance(distances, ...)	Test if distances are significantly different from the null distribution
tl.predict_knn(distances, y_true[, ...])	Predict values of y_true using K-nearest neighbors
tl.evaluate_prediction(y_true, y_pred, task, ...)	Evaluate how well y_pred predicts y_true
tl.test_proportions(target, groups)	Run statistical test to check if distribution of target differs between groups
tl.evaluate_representation(distances, target)	Evaluate representation of target for the given distance matrix