splicejac.tools.grn_comparison

functions to quantify grn similarity across cell states

Module Contents

Functions

edge_detection_score(m1, m2)

Compute AUROC/AUPRC related metrics between matrices m1 and m2 disregarding the interaction signs

sign_detection_score(m1, m2)

Compute AUROC/AUPRC related metrics between matrices m1 and m2 when considering the interaction signs

grn_comparison(adata)

Compute AUROC/AUPRC scores for all pairs of state-specific gene regulatory networks

splicejac.tools.grn_comparison.edge_detection_score(m1, m2)

Compute AUROC/AUPRC related metrics between matrices m1 and m2 disregarding the interaction signs

m1: ~numpy.ndarray

observation matrix

m2: ~numpy.ndarray

ground truth matrix

fpr: false positive rate tpr: true positive rate auroc: area under the receiver characteristic curve precision: precision recall: recall auprc: area under the precision recall curve

splicejac.tools.grn_comparison.sign_detection_score(m1, m2)

Compute AUROC/AUPRC related metrics between matrices m1 and m2 when considering the interaction signs

m1: observation matrix m2: ground truth matrix

fpr: float

false positive rate

tpr: float

true positive rate

auroc: float

area under the receiver characteristic curve

precision: float

precision

recall: float

recall

auprc: float

area under the precision recall curve

splicejac.tools.grn_comparison.grn_comparison(adata)

Compute AUROC/AUPRC scores for all pairs of state-specific gene regulatory networks Results are stored in adata.uns[‘comparison_scores’]

adata: ~anndata.AnnData

count matrix

None