0.14.2 (24-11-2025)

• remove direct link dependencies to allow PyPI upload

0.14.1 (21-11-2025)

• pin statsmodels to >= 0.14.5

0.14.0 (21-11-2025)

• fix dimension reduction and neighbor graph overwritten by batch correction (#343)
• remove deprecated define_pc (replaced by propose_pcs)
• implemented dim_red
• fix docker image build (#400)
• connectionPlot: fix legend title issues (#369) and add parameters to edit the ranges displayed in the legends
• hairball plot: add colormap parameter
• cyclone plot: add restrict_to parameter (#314), fix color_max plot/colorbar mismatch (#373)
• fix sctoolbox.plotting.qc_filter.quality_violin warning (#347)
• fix sctoolbox.plotting.marker_genes.rank_genes_plot, "n_genes" and "var_names" are mutually exclusive (#315)
• sctoolbox.tools.download_data: s3_downloader, download_data add option to display progress bar
• add functions to download exemplary RNA and ATAC datasets (#376)
• implemented sctoolbox.utils.adata.tidy_layers
• prepare_atac_anndata: remove "path_h5ad" parameter (#335)
• fix pl.marker_genes.rank_genes_plot, "n_genes" and "var_names" are mutually exclusive (#315)
• fix tools.marker_genes.get_rank_genes_tables may overwrite sheets (#332)
• automate PyPi release
• implemented sctoolbox.utils.adata.tidy_layers (#363)
• fix error making assembly notebook mandatory when generating report (#383)
• replace "/" in AnnData before saving to avoid errors (#388)
• CI/CD split notebook job into ATAC and RNA
• sctoolbox.utils.bioutils._bed_is_sorted and _read_bedfile account for comment lines.
• migrate to numpy 2
• Fix error in bam_adata_ov if some reads do not contain the tag
• add advanced tutorial notebooks (!448)
• Add function from_R (#395)
• update analysis reports (#225, #392)

Changes to notebooks

• temporarily remove the velocity notebook from testing (#397)
• update proportion_analysis notebook to be compatible with scanpro 0.4.0
• move gene annotation (clustering -> QC; #357)
• move HFV (QC -> normalization; #355)
• ATAC & RNA assembly: add section to adjust layers (#363)
• Allow multiple .rds/r.obj files to be read and merged into one adata (#395)
• renamed 0A2_ligand_receptor_differences.ipynb to 0A2_ligand_receptor_hub_genes.ipynb

0.13.5 (20-08-2025)

• fix pydeseq2 bug

0.13.4 (19-08-2025)

• implemented global thread settings
• fixed spelling mistake in title of predict cell cycle plot (#382)

0.13.3 (15-08-25)

• from_single_mtx: auto-adjust if the var/ obs file is one line of from the expected mtx
• add rasterization to reduce resource demand for big figures (plot_embedding, feature_per_group, search_umap-/ search_tnse_parameters, anndata_overview, pairwise_scatter)

0.13.2 (14-08-25)

• set bbknn(computation="cKDTree") and scanorama(approx=False) to fix issue with annoy package and processor architecture

0.13.1 (13-08-25)

• scrublet: use forkserver to start separte processes to enable multiprocessing (#380)
• enable AnnData write compression (this was default prior to AnnData version 0.6.16)

0.13.0 (08-08-25)

• predict_cell_cycle: implemented "gene_column" parameter
• restrict the maximum size of a figure (2^16 pixle)
• from_h5ad: expose concadata "label" parameter
• Add delete_obs and delete_var parameters to prepare_for_cellxgene() (#287)
• Add ability to use a peaks BED file to assemble var of ATAC-data (#346)
• set scanpy>=1.11 to fix run_rank_genes error (#331)
• GSEA: Revise gsea analysis: Bugfixes, save results into adata, rework plots (#345)
• Add column exists check to bioutils.pseudobulk_table (#356)
• Fix main title in receptor-ligand network plot (#358)
• load_h5ad(): warn if adata.raw is found. (#352)
• receptor-ligand: nan zscore to 0 (#302)
• adjust to altered scanpy.normalize_total behavior (#370)
• Add function to download tutorial data
• plot_pca_variance: add selected variance line; allow log-scale
• add global dpi setting
• implemented suppress_logging, get_version_report, plot_table, update_yml, generate_report
• Allow differential R-L plots to be asved as PDF
• receptor-ligand: adjust minimum line width in connectionPlot
• Added new notebook testdata and references. (partly #338)
• Improved flexibility in adata creation from mtx. (#365)
• lsi: fixed bug scaling the total variance explained to 100%

Changes to notebooks

• velocity: Changed scvelo.read() to scanpy.read() in the velocity notebook due to deprecation (#344)
• General: prepare_for_cellxgene: Set mampok version to 3.0.6
• General: prepare_for_cellxgene: Add metadata parameter for mamplan correlation
• General: prepare_for_cellxgene: Add delete column option for .obs and .var
• RNA 02 QC: fixed bug causing initial var thresholds to be ignored
• General: pseudotime_analysis: color dendrogram for clustering instead of segment
• add layer option to notebooks that utilize the matrix (#342)
• pptreport integration:
  • 01-RNA
  • 02-RNA
  • 03-RNA
  • 04-RNA
  • 0A1-RNA receptor-ligand
  • 0A2-RNA receptor-ligand differences
  • 0B-RNA velocity notebook
  • general group_markers
  • general pseudotime
  • general proportion
  • general GSEA
  • general annotation
  • 01-ATAC
  • 02-ATAC
  • 03-ATAC
  • 04-ATAC
• RNA: implemented report notebook
• General: annotation: add min_hits parameter
• RNA/ ATAC 03: allow to choose the number of computed PCs
• RNA/ ATAC 01: allow to choose batch name
• General: Move settings to config file
• velocity: Add missing save/embedding options
• atac_analysis: assembling: Changed to new testdata.
• ATAC: Implemented TOBIAS footprinting notebook
• General: Pseudotime: Remove threads parameter from dendrogram function

0.12.0 (19-12-24)

• add contrasts parameter to tools.marker_genes.run_deseq2
• tools.marker_genes.pairwise_rank_genes check minimum amount of groups

Changes to notebooks

• General: group_marker: set n_genes as top_n for get_rank_genes_tables
• ATAC 01 assembly: fix error when selecting the path_mtx (3rd) option (#326)
• General: prepare_for_cellxgene: set required mampok version to 3.0.5