omicverse.pl.dynamic_heatmap

omicverse.pl.dynamic_heatmap(adata, var_names, pseudotime, *, lineage_key=None, lineages=None, max_lineages=2, top_features=None, cell_bins=100, use_cell_columns=True, use_fitted=True, aggregate='mean', smooth_window=15, score_smooth_window=21, fitted_window=31, figsize=(8, 6), layer=None, use_raw=False, gene_symbols=None, standard_scale='var', cmap='viridis', pseudotime_cmap='cividis', order_by='peak', reverse_ht=None, n_split=None, split_method='kmeans-peaktime', cluster_features_by=None, row_cluster=False, col_cluster=False, show_row_names=False, show_column_names=False, pseudotime_label=None, cell_annotation=None, separate_annotation=None, separate_annotation_type='auto', separate_smooth_window=21, feature_labels=None, top_label_features=10, legend=True, legend_style='tight', border=False, save=False, save_pathway='', show=False)[source]

Plot dynamic feature trends along pseudotime, optionally by lineage.

Parameters:

  • adata (AnnData) – Annotated data matrix containing the expression values.

  • var_names – Features to plot. Accepts a list of genes, a single gene name, a mapping of group label to gene list, or None to start from all available features.

  • pseudotime (str) – Key in adata.obs containing the continuous ordering variable.

  • lineage_key (str (default: None)) – Optional key in adata.obs describing lineage membership.

  • lineages (default: None) – Optional lineage subset to include.

  • max_lineages (int (default: 2)) – Maximum number of lineages retained after ordering.

  • top_features (int (default: None)) – Keep only the highest-scoring dynamic features after ranking.

  • cell_bins (int (default: 100)) – Number of bins used when aggregating cells along pseudotime.

  • use_cell_columns (bool (default: True)) – Whether to keep columns at cell/bin resolution instead of grouped lineage blocks.

  • use_fitted (bool (default: True)) – Whether to smooth/fitted the matrix before plotting.

  • aggregate (str (default: 'mean')) – Aggregation function used when binning cells, for example 'mean'.

  • smooth_window (int (default: 15)) – Window sizes used for smoothing, feature scoring, and fitted trends.

  • score_smooth_window (int (default: 21)) – Window sizes used for smoothing, feature scoring, and fitted trends.

  • fitted_window (int (default: 31)) – Window sizes used for smoothing, feature scoring, and fitted trends.

  • figsize (tuple (default: (8, 6))) – Figure size (width, height) in inches.

  • layer (str (default: None)) – Expression source selection passed through to obs_df.

  • use_raw (bool (default: False)) – Expression source selection passed through to obs_df.

  • gene_symbols (str (default: None)) – Expression source selection passed through to obs_df.

  • standard_scale (str (default: 'var')) – Scaling mode. Supported values include 'var', 'zscore', 'group', 'minmax', 'obs', 'raw', or None.

  • cmap (str (default: 'viridis')) – Colormaps for the heatmap body and pseudotime annotations.

  • pseudotime_cmap (str (default: 'cividis')) – Colormaps for the heatmap body and pseudotime annotations.

  • order_by (str (default: 'peak')) – Feature ordering strategy, e.g. 'peak' or 'valley'.

  • reverse_ht (default: None) – Optional lineage names or positions whose pseudotime direction should be reversed before plotting.

  • n_split (int (default: None)) – Controls for splitting dynamic features into blocks.

  • split_method (str (default: 'kmeans-peaktime')) – Controls for splitting dynamic features into blocks.

  • cluster_features_by (str (default: None)) – Optional metadata-driven feature clustering mode.

  • row_cluster (bool (default: False)) – Whether to add hierarchical dendrograms for rows or columns.

  • col_cluster (bool (default: False)) – Whether to add hierarchical dendrograms for rows or columns.

  • show_row_names (bool (default: False)) – Whether to show row or column labels directly on the plot.

  • show_column_names (bool (default: False)) – Whether to show row or column labels directly on the plot.

  • pseudotime_label (default: None) – Optional custom label for the pseudotime annotation.

  • cell_annotation (default: None) – Optional adata.obs key or keys to annotate cells/bins above the heatmap.

  • separate_annotation (default: None) – Controls for auxiliary smoothed annotation tracks.

  • separate_annotation_type (str (default: 'auto')) – Controls for auxiliary smoothed annotation tracks.

  • separate_smooth_window (int (default: 21)) – Controls for auxiliary smoothed annotation tracks.

  • feature_labels (default: None) – Explicit feature labels to force-display.

  • top_label_features (int (default: 10)) – Number of automatically selected feature labels to display when feature_labels is not provided.

  • legend (bool (default: True)) – Legend visibility and layout strategy.

  • legend_style (str (default: 'tight')) – Legend visibility and layout strategy.

  • border (bool (default: False)) – Whether to draw a visible border around heatmap axes after rendering.

  • save (bool | str (default: False)) – Output controls. save may be True/False or a concrete path.

  • save_pathway (str (default: '')) – Output controls. save may be True/False or a concrete path.

  • show (bool (default: False)) – Output controls. save may be True/False or a concrete path.

Returns:

Returns the assembled Marsilea board by default, or the rendered figure when rendering is triggered for saving/showing/tight legends.

Return type:

marsilea.StackBoard or matplotlib.figure.Figure