PySCENIC Versions Save

pySCENIC is a lightning-fast python implementation of the SCENIC pipeline (Single-Cell rEgulatory Network Inference and Clustering) which enables biologists to infer transcription factors, gene regulatory networks and cell types from single-cell RNA-seq data.

0.9.19

4 years ago

-BugFixes: #100 and #51.

0.9.18

4 years ago

In the modules_from_adjacencies function, the default value of rho_mask_dropouts is changed to False. This now matches the behavior of the R version of SCENIC. Since this is likely to change the final output regulons slightly, warnings have been added to the logging system. The cli version has an additional option to turn dropout masking back on (--mask_dropouts).

0.9.17

4 years ago

In the modules_from_adjacencies function, the default value of rho_mask_dropouts is changed to False. This now matches the behavior of the R version of SCENIC. Since this is likely to change the final output regulons slightly, warnings have been added to the logging system. The cli version has an additional option to turn dropout masking back on (--mask_dropouts).

0.9.16

4 years ago

BugFix: minor bugfix in calculation of Regulon Specificity Score (RSS).

0.9.15

4 years ago

New features: multiprocessing support for the binarization function.

0.9.14

4 years ago
  • BugFix: issue #87 - Binarization fails with regulon of few genes.
  • BugFix: Always strip off extensions of filenames with os.path.splitext()[0] instead of using split(".")[0], which wouldn't work properly for filenames which contain embedded dots.

0.9.13

4 years ago
  • New functionality: Support for Regulon Specificity Score (RSS) and plotting functions for AUCell distributions and RSS values for a cell type.
  • Bug Fixes: #70 and #81

0.9.11

4 years ago
  • New functionality: new algorithm to define binarization threshold on AUC values of a regulon. Hartigan's Dip Test (HDT) is used to decide if the distribution of AUC values deviates from unimodality. If this is the case, a bimodal gaussian mixture model is fit to capture the two modes of the distribution. The trough between these two modes is the threshold and is derived by minimization on the kernel density smoothed histogram.

0.9.10

4 years ago

BugFix: CLI grn - output of list of adjacencies is TSV or CSV (default) based on provided file extension.

0.9.9

5 years ago

New functionality: support for multiple embeddings then exporting to loom file.