omicverse.single.pyVIA

omicverse.single.pyVIA#

class omicverse.single.pyVIA(adata, adata_key='X_pca', adata_ncomps=80, basis='X_umap', clusters='', dist_std_local=2, jac_std_global=0.15, labels=None, keep_all_local_dist='auto', too_big_factor=0.4, resolution_parameter=1.0, partition_type='ModularityVP', small_pop=10, jac_weighted_edges=True, knn=30, n_iter_leiden=5, random_seed=42, num_threads=-1, distance='l2', time_smallpop=15, super_cluster_labels=False, super_node_degree_list=False, super_terminal_cells=False, x_lazy=0.95, alpha_teleport=0.99, root_user=None, preserve_disconnected=True, dataset='', super_terminal_clusters=[], is_coarse=True, csr_full_graph='', csr_array_locally_pruned='', ig_full_graph='', full_neighbor_array='', full_distance_array='', df_annot=None, preserve_disconnected_after_pruning=False, secondary_annotations=None, pseudotime_threshold_TS=30, cluster_graph_pruning_std=0.15, visual_cluster_graph_pruning=0.15, neighboring_terminal_states_threshold=3, num_mcmc_simulations=1300, piegraph_arrow_head_width=0.1, piegraph_edgeweight_scalingfactor=1.5, max_visual_outgoing_edges=2, via_coarse=None, velocity_matrix=None, gene_matrix=None, velo_weight=0.5, edgebundle_pruning=None, A_velo=None, CSM=None, edgebundle_pruning_twice=False, pca_loadings=None, time_series=False, time_series_labels=None, knn_sequential=10, knn_sequential_reverse=0, t_diff_step=1, single_cell_transition_matrix=None, embedding_type='via-mds', do_compute_embedding=False, color_dict=None, user_defined_terminal_cell=[], user_defined_terminal_group=[], do_gaussian_kernel_edgeweights=False, RW2_mode=False, working_dir_fp='/home/shobi/Trajectory/Datasets/')[source]#

VIA-based trajectory inference and lineage visualization wrapper.

Parameters:

  • adata (AnnData)

  • adata_key (str (default: 'X_pca'))

  • adata_ncomps (int (default: 80))

  • basis (str (default: 'X_umap'))

  • clusters (str (default: ''))

  • dist_std_local (float (default: 2))

  • labels (Optional[ndarray] (default: None))

  • too_big_factor (float (default: 0.4))

  • resolution_parameter (float (default: 1.0))

  • partition_type (str (default: 'ModularityVP'))

  • small_pop (int (default: 10))

  • jac_weighted_edges (bool (default: True))

  • knn (int (default: 30))

  • n_iter_leiden (int (default: 5))

  • random_seed (int (default: 42))

  • super_cluster_labels (bool (default: False))

  • super_node_degree_list (bool (default: False))

  • super_terminal_cells (bool (default: False))

  • x_lazy (float (default: 0.95))

  • alpha_teleport (float (default: 0.99))

  • preserve_disconnected (bool (default: True))

  • dataset (str (default: ''))

  • super_terminal_clusters (list (default: []))

  • csr_full_graph (ndarray (default: ''))

  • preserve_disconnected_after_pruning (bool (default: False))

  • secondary_annotations (Optional[list] (default: None))

  • pseudotime_threshold_TS (int (default: 30))

  • cluster_graph_pruning_std (float (default: 0.15))

  • visual_cluster_graph_pruning (float (default: 0.15))

  • max_visual_outgoing_edges (int (default: 2))

  • time_series_labels (Optional[list] (default: None))

  • knn_sequential (int (default: 10))

  • knn_sequential_reverse (int (default: 0))

  • t_diff_step (int (default: 1))

  • embedding_type (str (default: 'via-mds'))

  • do_compute_embedding (bool (default: False))

  • color_dict (Optional[dict] (default: None))

  • user_defined_terminal_cell (list (default: []))

  • user_defined_terminal_group (list (default: []))

  • do_gaussian_kernel_edgeweights (bool (default: False))

  • RW2_mode (bool (default: False))

  • working_dir_fp (str (default: '/home/shobi/Trajectory/Datasets/'))

__init__(adata, adata_key='X_pca', adata_ncomps=80, basis='X_umap', clusters='', dist_std_local=2, jac_std_global=0.15, labels=None, keep_all_local_dist='auto', too_big_factor=0.4, resolution_parameter=1.0, partition_type='ModularityVP', small_pop=10, jac_weighted_edges=True, knn=30, n_iter_leiden=5, random_seed=42, num_threads=-1, distance='l2', time_smallpop=15, super_cluster_labels=False, super_node_degree_list=False, super_terminal_cells=False, x_lazy=0.95, alpha_teleport=0.99, root_user=None, preserve_disconnected=True, dataset='', super_terminal_clusters=[], is_coarse=True, csr_full_graph='', csr_array_locally_pruned='', ig_full_graph='', full_neighbor_array='', full_distance_array='', df_annot=None, preserve_disconnected_after_pruning=False, secondary_annotations=None, pseudotime_threshold_TS=30, cluster_graph_pruning_std=0.15, visual_cluster_graph_pruning=0.15, neighboring_terminal_states_threshold=3, num_mcmc_simulations=1300, piegraph_arrow_head_width=0.1, piegraph_edgeweight_scalingfactor=1.5, max_visual_outgoing_edges=2, via_coarse=None, velocity_matrix=None, gene_matrix=None, velo_weight=0.5, edgebundle_pruning=None, A_velo=None, CSM=None, edgebundle_pruning_twice=False, pca_loadings=None, time_series=False, time_series_labels=None, knn_sequential=10, knn_sequential_reverse=0, t_diff_step=1, single_cell_transition_matrix=None, embedding_type='via-mds', do_compute_embedding=False, color_dict=None, user_defined_terminal_cell=[], user_defined_terminal_group=[], do_gaussian_kernel_edgeweights=False, RW2_mode=False, working_dir_fp='/home/shobi/Trajectory/Datasets/')[source]#

Initialize a pyVIA trajectory inference model.

Parameters:

  • adata (anndata.AnnData) – Input single-cell AnnData object.

  • adata_key (str, default='X_pca') – Key in adata.obsm used as low-dimensional input for VIA graph construction.

  • adata_ncomps (int, default=80) – Number of components retained from adata.obsm[adata_key].

  • basis (str, default='X_umap') – Embedding key in adata.obsm used for plotting and trajectory overlays.

  • clusters (str, default='') – Column name in adata.obs storing initial cluster/cell-type labels.

  • dist_std_local (float, default=2) – Local pruning strength for PARC/VIA graph edges.

  • jac_std_global (float, default=0.15) – Global Jaccard-based pruning threshold.

  • labels (numpy.ndarray, optional) – Optional external labels used instead of adata.obs[clusters].

  • keep_all_local_dist (str or bool, default='auto') – Whether to keep all local distances before pruning.

  • too_big_factor (float, default=0.4) – Re-partition clusters larger than this fraction of total cells.

  • resolution_parameter (float, default=1.0) – Leiden/PARC partition resolution.

  • partition_type (str, default='ModularityVP') – Graph partition strategy passed to PARC/VIA.

  • small_pop (int, default=10) – Minimum cluster size considered stable.

  • jac_weighted_edges (bool, default=True) – Whether to weight graph edges by Jaccard overlap.

  • knn (int, default=30) – Number of nearest neighbors used to build the cell graph.

  • n_iter_leiden (int, default=5) – Number of Leiden refinement iterations.

  • random_seed (int, default=42) – Random seed for reproducible graph partitioning.

  • num_threads (int, default=-1) – Number of CPU threads; -1 lets backend decide automatically.

  • distance (str, default='l2') – Distance metric used in neighbor search.

  • time_smallpop (int, default=15) – Iteration/time control for handling small populations.

  • super_cluster_labels (bool, default=False) – Whether to compute super-cluster labels on top of base clusters.

  • super_node_degree_list (bool, default=False) – Whether to expose super-node degree statistics.

  • super_terminal_cells (bool, default=False) – Whether to infer terminal states at super-cluster granularity.

  • x_lazy (float, default=0.95) – Lazy random-walk parameter controlling self-transition probability.

  • alpha_teleport (float, default=0.99) – Teleportation probability for Markov diffusion.

  • root_user (list, optional) – User-defined root cell indices or root groups for pseudotime orientation.

  • preserve_disconnected (bool, default=True) – Keep disconnected graph components during trajectory construction.

  • dataset (str, default='') – Dataset/root mode flag used by VIA internals.

  • super_terminal_clusters (list, default=[]) – User-provided terminal super-cluster IDs.

  • is_coarse (bool, default=True) – Whether current model is coarse stage in a coarse-to-fine workflow.

  • csr_full_graph (numpy.ndarray or scipy.sparse matrix, optional) – Optional precomputed full graph adjacency.

  • csr_array_locally_pruned (numpy.ndarray or scipy.sparse matrix, optional) – Optional precomputed locally pruned graph.

  • ig_full_graph (igraph.Graph, optional) – Optional pre-built igraph object.

  • full_neighbor_array (numpy.ndarray, optional) – Optional precomputed neighbor index array.

  • full_distance_array (numpy.ndarray, optional) – Optional precomputed neighbor distance array.

  • df_annot (pandas.DataFrame, optional) – Cell annotation table used by VIA plotting and summaries.

  • preserve_disconnected_after_pruning (bool, default=False) – Whether to retain disconnected components after graph pruning.

  • secondary_annotations (list, optional) – Additional annotation tracks for visualization.

  • pseudotime_threshold_TS (int, default=30) – Threshold used in terminal-state calling for time-series mode.

  • cluster_graph_pruning_std (float, default=0.15) – Pruning strength for cluster-level graph.

  • visual_cluster_graph_pruning (float, default=0.15) – Additional pruning for visualized cluster graph.

  • neighboring_terminal_states_threshold (int, default=3) – Merge threshold for nearby terminal states.

  • num_mcmc_simulations (int, default=1300) – Number of random-walk/MCMC simulations for lineage probabilities.

  • piegraph_arrow_head_width (float, default=0.1) – Arrow head width in pie-chart trajectory graph.

  • piegraph_edgeweight_scalingfactor (float, default=1.5) – Edge-width scaling in pie-chart trajectory graph.

  • max_visual_outgoing_edges (int, default=2) – Maximum outgoing edges per node in visual graph rendering.

  • via_coarse (object, optional) – Optional coarse VIA model used for hierarchical runs.

  • velocity_matrix (numpy.ndarray, optional) – RNA velocity matrix used to orient transitions.

  • gene_matrix (numpy.ndarray, optional) – Expression matrix aligned with velocity_matrix.

  • velo_weight (float, default=0.5) – Relative weight of velocity-informed transitions.

  • edgebundle_pruning (float, optional) – Pruning threshold before edge bundling.

  • A_velo (numpy.ndarray, optional) – Cluster-level velocity transition matrix.

  • CSM (Any, optional) – Optional custom transition/similarity matrix.

  • edgebundle_pruning_twice (bool, default=False) – Whether to apply two-pass pruning before edge bundling.

  • pca_loadings (numpy.ndarray, optional) – PCA loadings used in velocity projection.

  • time_series (bool, default=False) – Whether to enable time-series constraints.

  • time_series_labels (list, optional) – Ordered temporal labels per cell.

  • knn_sequential (int, default=10) – Number of forward temporal neighbors (t to t+1).

  • knn_sequential_reverse (int, default=0) – Number of reverse temporal neighbors (t to t-1).

  • t_diff_step (int, default=1) – Maximum temporal step size allowed in sequential KNN links.

  • single_cell_transition_matrix (Any, optional) – Optional precomputed single-cell transition matrix.

  • embedding_type (str, default='via-mds') – Embedding algorithm used by VIA (for example 'via-mds').

  • do_compute_embedding (bool, default=False) – Whether to compute VIA embedding during initialization.

  • color_dict (dict, optional) – Mapping from category to plotting color.

  • user_defined_terminal_cell (list, default=[]) – User-defined terminal cell indices.

  • user_defined_terminal_group (list, default=[]) – User-defined terminal cluster/group labels.

  • do_gaussian_kernel_edgeweights (bool, default=False) – Whether to use Gaussian-kernel edge weights.

  • RW2_mode (bool, default=False) – Enable RW2 random-walk mode in VIA.

  • working_dir_fp (str, default='/home/shobi/Trajectory/Datasets/') – Working directory used by VIA for intermediate files.

Methods

__init__(adata[, adata_key, adata_ncomps, ...])

Initialize a pyVIA trajectory inference model.

get_piechart_dict([label, clusters])

Get cluster composition dictionary for pie chart visualization.

get_pseudotime([adata])

Extract the pseudotime values computed by VIA.

plot_clustergraph(gene_list[, arrow_head, ...])

Plot cluster graph with aggregated gene-expression nodes.

plot_gene_trend([gene_list, figsize, ...])

Plot smoothed gene trends along VIA pseudotime.

plot_gene_trend_heatmap(gene_list[, ...])

Plot lineage-specific heatmaps of gene trends.

plot_lineage_probability([clusters, basis, ...])

Plot lineage membership probabilities in embedding space.

plot_piechart_graph([clusters, type_data, ...])

Plot two subplots with clustergraph representation showing cluster composition and pseudotime/gene expression.

plot_stream([clusters, basis, density_grid, ...])

Plot streamlines of inferred cell-state flow on embedding.

plot_trajectory_gams([clusters, basis, ...])

Project coarse VIA trajectories onto embedding.

run()

Calculate the VIA graph and pseudotime.
Contents