splicejac.tools.grn_statistics
evaluate gene statistics of the state-specific GRNs
Module Contents
Functions
|
Compute the signaling scores of all genes in each cell states |
|
Compute statistical metrics including standard deviation, range, and interquartile range of a 1D vector |
|
Compute state-specific GRN statistics including gene centrality, incoming, outgoing and total signaling strength |
|
Computes various statistics on the cell state specific GRNs. The statistics are added to adata.uns['GRN_statistics'] |
- splicejac.tools.grn_statistics.signaling_score(adata)
Compute the signaling scores of all genes in each cell states
Scores are defined based on number of incoming/outgoing edges and weighted sum of incoming/outgoing edges Results are saved in adata.uns[‘signaling_scores’]
- adata: ~anndata.AnnData
count matrix
None
- splicejac.tools.grn_statistics.compute_metrics(dat)
Compute statistical metrics including standard deviation, range, and interquartile range of a 1D vector
- dat: ~numpy.ndarray
1D array of measurements
- metrics_df: ~pandas.dataFrame
dataframe of metrics
- splicejac.tools.grn_statistics.GRN_cluster_variation(adata)
Compute state-specific GRN statistics including gene centrality, incoming, outgoing and total signaling strength Results are stored in adata.uns[‘cluster_variation’] and adata.uns[‘cluster_average’]
- adata: ~anndata.AnnData
count matrix
None
- splicejac.tools.grn_statistics.grn_statistics(adata, weight_quantile=0.5, k=None, normalized=True, weight=None, endpoints=False, seed=None)
Computes various statistics on the cell state specific GRNs. The statistics are added to adata.uns[‘GRN_statistics’]
For a more detailed discussion of several of the parameters, please see the betweenness_centrality from Networkx (https://networkx.org/documentation/networkx-1.10/reference/generated/networkx.algorithms.centrality.betweenness_centrality.html) Results are stored in adata.uns[‘GRN_statistics’], adata.uns[‘cluster_variation’] and adata.uns[‘cluster_average’]
- adata: ~anndata.AnnData
count matrix
- weight_quantile: float (default: 0.5)
cutoff for weak gene-gene interactions between 0 and 1
- k: int or None (default=`None`)
number of nodes considered to compute betweenness centrality. k=None implies that all edges are used
- normalized: Bool (default: True)
if True, betweenness centrality values are normalized
- weight: str (default: None)
If None, all edge weights are considered equal
- endpoints: Bool (default: False)
If True, include the endpoints in the shortest path counts during betweenness centrality calculation
- seed: int (default: None)
seed for betweenness centrality calculation
None