U-Chrom¶

Universal Chromatin: one toolkit for 3-D genome data from sequencing and imaging.

Single-cell Hi-C, spatial Hi-C, bulk Hi-C, chromatin tracing and DNA seqFISH+ all describe the same thing — where loci sit relative to each other, cell by cell — yet each comes with its own formats and tools. U-Chrom puts them in one data model, ChromData (cells › traces › spots on a shared locus axis, with contact maps, RNA, chromatin marks and cell positions linked per cell), and builds the analyses on it: import, 3-D reconstruction, structure calling, per-locus features, single-cell embeddings, an interactive web browser and an open data atlas.

import uchrom as uc
import uchrom.datasets as ds
from chromdata import ChromData

cd = ChromData.from_fofct(ds.fetch("takei"))           # chromatin tracing (4DN FOF-CT), fetched once
tads = uc.tl.call_tads(cd, chrom="chr3")                # TADs from 3-D distances
cd.write("takei2021.chromdata.zarr")                    # one store, read in parts later

python -m uchrom_browser takei2021.chromdata.zarr       # look at it in the web browser

New here? Start with installation, the quick start and the concepts behind the data model.

By data type¶

Each chapter takes one kind of data from its files to a ChromData and through what is specific to it, with tutorials on real data.

Data	What the chapter covers
Chromatin tracing	FOF-CT, PyHiM and other tables; raw detections into traces; filling missing loci
Imaging multi-omics: DNA seqFISH+	DNA positions with chromatin marks, RNA and cell positions in the same cells
Single-cell and spatial Hi-C	contacts per cell, linked contact maps, 3-D structures of single cells (NucDynamics, EMber), embeddings from contacts
Bulk Hi-C	contact maps to 3-D: one consensus structure (MDS), a population of structures (IGM), Hi-C with tracing (GEM-FISH)

Common to all data¶

Chapter	What it covers	Package / module
Data model & storage	`ChromData`, the `.chromdata.zarr` store, backed and remote reads, linked and embedded modalities	`chromdata`
Structure calling	loops, TADs, domains per allele, A/B compartments from 3-D distances	`uchrom.strc`
Per-locus features	distance maps, variance features, signals and peaks along the genome	`uchrom.fea`
Cells & embeddings	embeddings and clusters of cells from RNA, chromatin marks, accessibility and contacts	`uchrom.emb`
Visualisation	the web browser, the data atlas, plots	`uchrom-browser`, `uchrom.pl`