chrysalis-st 0.1.1

Powerful and lightweight package to identify tissue compartments in spatial transcriptomics datasets.

chrysalis is a spatial domain detection and visualization tool that generates neat and meaningful visual representations of spatial transcriptomics datasets. It achieves this by leveraging PCA, archetypal analysis and spatially variable gene detection. Moreover, it seamlessly integrates into scanpy based pipelines.

By combining PCA with AA, chrysalis can define distinct tissue compartments and cellular niches, which can be highlighted with specific colors. For instance, in the V1_Human_Lymph_Node dataset, chrysalis can identify and highlight various regions, such as germinal centers (yellow), B cell follicles (dark orange), and T cell compartments (lime). You can find more examples in the gallery section.

Package

chrysalis can be used with any pre-existing anndata snapshot of 10X Visium, Slide-seqV2 and Stereo-seq datasets generated with scanpy, and on new samples without the need of preprocessing. It is designed to be as lightweight as possible, however currently it relies on libpysal for its fast implementation of Moran's I.

chrysalis requires the following packages:

• numpy
• pandas
• matplotlib
• scanpy
• pysal
• archetypes
• scikit_learn
• scipy
• tqdm
• seaborn

To install chrysalis:

pip install chrysalis-st

Usage

import chrysalis as ch
import scanpy as sc
import matplotlib.pyplot as plt

adata = sc.datasets.visium_sge(sample_id='V1_Human_Lymph_Node')

sc.pp.calculate_qc_metrics(adata, inplace=True)
sc.pp.filter_cells(adata, min_counts=6000)
sc.pp.filter_genes(adata, min_cells=10)

ch.detect_svgs(adata)

sc.pp.normalize_total(adata, inplace=True)
sc.pp.log1p(adata)

ch.pca(adata)

ch.aa(adata, n_pcs=20, n_archetypes=8)

ch.plot(adata)
plt.show()

Documentation and API details

User documentation is available at: https://chrysalis.readthedocs.io/

Gallery