Version 0.2.3 Feb 13, 2021¶
- tl.recover_dynamics: Multicore implementation thanks to M Klein, Y Schaelte, P Weiler
- CI now runs on GitHub Actions
New utility functions:
- utils.gene_info: Retrieve gene information from biothings client.
- utils.convert_to_ensembl and utils.convert_to_gene_names: Converting ensembl IDs into gene names and vice versa.
Version 0.2.2 July 22, 2020¶
- tl.paga: PAGA graph with velocity-directed edges.
- Black code style
Version 0.2.1 May 28, 2020¶
- Correct identification of root cells in tl.latent_time thanks to M Lange
- Correct usage of latent_time prior in tl.paga thanks to G Lubatti
Version 0.2.0 May 12, 2020¶
- RNA velocity basics
- Dynamical Modeling
- Differential Kinetics
- tl.differential_kinetic_test: introduced a statistical test to detect different kinetic regimes.
- tl.rank_dynamical_genes: introduced a gene ranking by cluster-wise likelihoods.
- tl.paga: introduced directed PAGA graph
- pl.scatter enhancements: linear and polynomical fits, gradient coloring
- pl.proportions: Pie and bar chart of spliced/unspliced proprtions.
- GridSpec: multiplot environment.
Version 0.1.20 Sep 5, 2019¶
- tl.recover_dynamics: introduced a dynamical model inferring the full splicing kinetics, thereby identifying all kinetic rates of transcription, splicing and degradation.
- tl.recover_latent_time: infers a shared latent time across all genes based on the learned splicing dynamics.
- pl.scatter enhancements: multiplots, rugplot, linear and polynomial fits, density plots, etc.
- pl.heatmap: heatmap / clustermap of genes along time coordinate sorted by expression along dynamics.
- New attributes in pp.filter_genes: min_shared_counts and min_shared_genes.
- Added fast neighbor search method: Hierarchical Navigable Small World graphs (HNSW)
Version 0.1.14 Dec 7, 2018¶
- New attriutes arrow_length and arrow_size for flexible adjustment of embedded velocities.
- pl.velocity_graph: Scatter plot of embedding with cell-to-cell transition connectivities.
- pl.velocity_embedding_stream: Streamplot visualization of velocities.
- Improve visualization of embedded single cell velocities (autosize, colors etc.)
- tl.cell_fate: compute cell-specific terminal state likelihood
- New attribute approx=True in tl.velocity_graph to enable approximate graph computation by performing cosine correlations on PCA space.
- Automatically detect whether data is already preprocessed.
Version 0.1.11 Oct 27, 2018¶
- settings.set_figure_params(): adjust matplotlib defaults for beautified plots
- improved default point and arrow sizes; improved quiver autoscale
- enable direct plotting of
- tl.velocity_confidence: Added two confidence measures ‘velocity_confidence’ and ‘velocity_confidence_transition’.
- tl.rank_velocity_genes: Added functionality to rank genes for velocity characterizing groups using a t-test.
- New attribute perc in tl.velocity enables extreme quantile fit, e.g. set perc=95.
- New attribute groups in tl.velocity enables velocity estimation only on a subset of the data.
- Improved tl.transition_matrix by incorporating self-loops via self_transitions=True and state changes that have negative correlation with velocity (opposite direction) via use_negative_cosines=True
- utils.merge to merge to AnnData objects such as already existing AnnData and newly generated Loom File.
Version 0.1.8 Sep 12, 2018¶
- support saving plots as pdf, png etc.
- support multiple colors and layers
- quiver autoscaling for velocity plots
- attributes added: figsize and dpi
- filter_and_normalize() instead of recipe_velocity()
- normalization of layers is done automatically when computing moments
- terminal_states: computes root and end points via eigenvalue decomposition thanks to M Lange
Version 0.1.5 Sep 4, 2018¶
- Support writing loom files
- Support both dense and sparse layers
- Plotting bugfixes
- Added pp.recipe_velocity()
Version 0.1.2 Aug 21, 2018¶
First alpha release of scvelo.