Browsing the atlas

The U-Chrom atlas is a public collection of 3-D genome datasets — single-cell Hi-C, spatial Hi-C, imaging — each converted to one self-contained .chromdata.zarr store and served from object storage (https://uchrom-atlas-r2.u-science.org). A store opens without downloading it: only what you look at or analyse is read.

On the web

  • The atlas page, uchrom-atlas.u-science.org: every dataset with its study, organism, tissue, number of cells, modalities and a thumbnail, grouped by study; each card links to the paper, the source data and the one-file download.

  • The hosted web browser, uchrom-browser.u-science.org: open a dataset from the Open dialog → Atlas and look at it — cells in their tissue, structures in 3-D, contact maps, embeddings — with nothing installed. The same dialog lists the atlas in your own browser (python -m uchrom_browser, see the web browser).

From Python

ds.list_atlas() reads the atlas catalog (catalog.json) and returns one row per store:

import uchrom.datasets as ds

atlas = ds.list_atlas()
atlas[["title", "organism", "n_cells", "modalities", "size_mb", "downloaded"]]
atlas[atlas["modalities"].map(lambda m: "3-D coordinates" in m)]      # stores with 3-D structures

Column

Meaning

title, study, organism, tissue, assembly

what the dataset is (the study’s citation and source accession are in catalog.json)

n_cells, modalities

cells (spots for spatial data) and what the store holds: Hi-C per cell, Hi-C per spot, Hi-C bulk, RNA, ATAC per cell, 3-D coordinates, signals per spot, images, embeddings, …

size_mb, download_mb

the store, and its one-file copy (.cdz)

downloaded

the .cdz is already in your data directory (fetching)

url

the address ds.load and the web browser open

The full catalog — descriptions, papers, DOIs, licences, studies — is chromdata.catalog.fetch_catalog(url) (a dict); python -m uchrom.datasets atlas prints the short table.

Another atlas

An atlas is any folder or bucket of stores with a catalog.json: ds.list_atlas(root) and ds.atlas(id, root=root) read another one, python -m uchrom_browser --atlas URL_OR_FOLDER (or UCHROM_ATLAS) lists it in the web browser. How a dataset is added to the public atlas — recipes, embedded copies, the catalog, the page — is described in the data atlas.