tfmindi.pl.tsne_logos

Contents

tfmindi.pl.tsne_logos#

tfmindi.pl.tsne_logos(adata, patterns=None, color_by='cluster_dbd', logo_width=1.0, logo_height=0.8, alpha=0.2, s=2, ic_threshold=0.2, min_nucleotides=4, min_seqlets=10, show_cluster_labels=True, show_legend=True, gray_background=True, cmap='tab20', **kwargs)#

Visualize seqlet clusters in t-SNE space with optional sequence logos at centroids.

This function can display both basic t-SNE scatter plots and enhanced plots with sequence logos. When show_logos=False, it delegates to the basic tsne() function for fast exploration. When show_logos=True, it renders sequence logos at cluster centroids for publication-quality visualization.

Parameters:

  • adata (AnnData) – AnnData object with t-SNE coordinates and cluster assignments. Must contain adata.obsm[“X_tsne”] and adata.obs[“leiden”].

  • patterns (dict[str, Pattern] | None (default: None)) – Dictionary mapping cluster IDs to Pattern objects with PWMs. Required when show_logos=True, optional when show_logos=False.

  • color_by (str (default: 'cluster_dbd')) – Column in adata.obs to use for coloring points.

  • logo_width (float (default: 1.0)) – Width of sequence logos relative to plot coordinates.

  • logo_height (float (default: 0.8)) – Height of sequence logos relative to plot coordinates.

  • alpha (float (default: 0.2)) – Transparency of scatter points.

  • s (float (default: 2)) – Size of scatter points.

  • ic_threshold (float (default: 0.2)) – Information content threshold for logo trimming.

  • min_nucleotides (int (default: 4)) – Minimum number of nucleotides required after trimming to show logo. Patterns with fewer nucleotides will be skipped to avoid showing noisy logos.

  • min_seqlets (int (default: 10)) – Minimum number of seqlets required in a cluster to display its logo. Clusters with fewer seqlets will be skipped to avoid showing weak motifs from small clusters.

  • show_cluster_labels (bool (default: True)) – Whether to show cluster ID labels.

  • show_legend (bool (default: True)) – Whether to show the legend (default: True).

  • gray_background (bool (default: True)) – Whether to use gray background for all scatter points to improve logo readability. When True, all points are colored gray and no legend is shown.

  • cmap (str (default: 'tab20')) – Colormap name for categorical data (default: “tab20”). Any valid matplotlib colormap name (e.g., “viridis”, “plasma”, “Set1”).

  • **kwargs – Additional arguments passed to render_plot() for styling and display options. Common options include width, height, title, xlabel, ylabel, show, save_path.

Return type:

Figure | None

Returns:

matplotlib.Figure or None Figure with t-SNE plot and optional sequence logos, or None if show=True.

Examples

>>> import tfmindi as tm
>>> # After clustering and pattern creation - full plot with logos
>>> tm.pl.tsne_logos(adata, patterns, color_by="cluster_dbd", width=12, height=10)
>>>
>>> # Publication plot with custom styling
>>> tm.pl.tsne_logos(adata, patterns, width=30, height=30, title="My Plot", show=False, save_path="plot.png")
Contents