Cellestial

The Grammar of Graphics for single-cell omics.

Installation

pip install cellestial

Introduction

Cellestial is an interactive and a highly customizable Single-Cell & Spatial omics data visualization library. Built on top Lets-Plot, it offers a ggplot-like layered and modular approach offering high customizability and publication-ready figures.

Cellestial is highly integrated with scverse's AnnData with room for integration with any upcoming single-cell omics data type in the Python single-cell omics ecosystem.

Cellestial leverages the performance of Polars ensuring speed and scalability.

Cellestial is built-on top the following core design philosophies:

1. Modularity over abstraction
2. Predictability over flexibility
3. Explicity over implicity
4. Simplicity & Expressiveness

with reproduciblity, intiutiveness & ease-of-use, and, of course, beautiful publication-ready plots/charts/figures in mind.

More on Design Philosophy

To see the example Figures visit Cellestial Webpage.

Example Plots Grid

Usage

import cellestial as cl

Tooltips

umap = cl.umap(
    data,
    key="cell_type_lvl1",
    size=1,
    axis_type="arrow",
    alpha=0.6,
    tooltips=["cell_type_lvl1", "n_genes","pct_counts_in_top_50_genes"],
    legend_ondata=True,
    ondata_size=12,
    ondata_fontface="bold",
    ondata_family="mono",
    ondata_alpha=0.8,
)

Layered/Modular Approach

The ggplot-like API allows adding layers and customizing the plot.

# modify plot size and color palette
umap += ggsize(800, 600) + scale_color_hue()
# highlight clusters
umap + cl.cluster_outlines(dim,["B Cells","Erythroid"])

Zooming and Paning Options

TODO

Grid plots

Instead of singular function names (umap), multi/grid plots requires the plural (umaps),providing predictability which guarentees the reproducibility.

Which are valid for all dimensional subsets (expression,pca,umap, tsne).

cl.umaps(
    data,
    keys=["leiden", "HBD", "NEAT1", "IGKC"],
    ncol=2,
    size=1,
    color_high="red",
) + ggsize(900, 600)