A Multimodal Single-Cell Epigenomic and 3D Genome Atlas of the Human Basal Ganglia

1. Data Availability

1.1 Raw fastq files

The raw fastq files (plate-level, before demultiplexing) are available at:

• NEMO: https://assets.nemoarchive.org/dat-aw6czix;
• FTP site: https://data.nemoarchive.org/bican/grant/BICAN_Mul_PN_Human/salk_ecker/epigenome/nuclei/m3C_seq/human/demultiplexed_fastq/

1.2 cell-level fastq

Cell-level fastq files (after demultiplexing) can be downloaded from GEO with accession ID GSE320293

1.3 Spatial data

MERFISH transcriptomic data will be available on BIL soon

1.4 bigwig & hic files

bigwig and hic files are available for download on Basal Ganglia Epigenome Browser

1.5 Processed data

Processed data are available on the NCBI Gene Expression Omnibus (GEO) and Figshare

GEO: GSE320293

• Pseudobulk allc files (Group level)
• HiC contact files (single-cell level)

Figshare (uploading)

• Pseudobulk allc files (Subclass level CG+CH): folder Subclass.allc
• adata: folder adata
• DMG: folder DMG
• DMR: folder DMR
• Enriched motif and TF: folder motif
• Normalized compartment scores across cell types at subclass level: HiC/NormalizedCompartmentScores.tsv
• Diff domain doundary: HiC/diff_boundary.tsv
• Loop at Subclass and Group levels: folder HiC/Subclass.loop and HiC/Group.loop
• Enhancer-Promoter Links: supplementary_tables/TableS7.enhancer_promoter_links.tsv
• TF-Target gene pairs: supplementary_tables/TableS8.Subclass.GRN.xlsx
• Supplementary Tables (for the manuscript): folder supplementary_tables

2. Mapping Pipeline

If you prefer to process the data from fastq files, please run dumultiplexing and mapping using our mapping pipeline:

• Local (or Google Cloud) snakemake pipeline: https://github.com/DingWB/cemba_data
• Broad WDL pipeline: https://broadinstitute.github.io/warp/docs/Pipelines/snm3C/README

3. Code and jupyter notebooks for clustering, cell type annotations and downstream analysis

• Clustering: See clustering folder
• Integration between snm3C-seq & sc-RNA: See integration folder
• Call DMR: https://github.com/lhqing/ALLCools
• Motif enrichment: motif_enrichment folder
• ABC Model: ABC folder
• Spatial Transcriptomics Data Analysis: https://github.com/a3klein/BG_snm3c-seq_MERFISH
• LDSC: https://github.com/rosanwang/ldsc-wrapper

4. Data exploration & visualization

• (1) Basal Ganglia Epigenome Browser: https://basalganglia.epigenomes.net/
• (2) Human Basal Ganglia (HBG) Anno-J Networked Genome Browser: https://neomorph.salk.edu/hbg/hbg.php

5. Citation

Ding, W., Klein, A., Baez-Becerra, C. T., Rink, J. A., Bartlett, A., Zeng, Q., ... & Ecker, J. R. (2026). A Multimodal Single-Cell Epigenomic and 3D Genome Atlas of the Human Basal Ganglia. bioRxiv, 2026-02.

6. FAQ

How to normalize the raw methylation fraction?

import os,sys
import pandas as pd
import anndata

adata_path="BG.gene-CHN.h5ad" # example adata path
gene="DRD2" # an example gene
clip_norm_value=10
raw_adata = anndata.read_h5ad(os.path.expanduser(adata_path), backed='r')
adata = raw_adata[:, gene].to_memory()
raw_adata.file.close()
cols = adata.obs.columns.tolist()
if 'prior_mean' in cols:
    na_sum = adata.to_df().isna().sum().sum()
    adata.X = adata.X / adata.obs.prior_mean.values[:, None]  # range = [0,1,10]
    if not clip_norm_value is None:
        if issparse(adata.X):
            X=adata.X.toarray()
        else:
            X=adata.X
        adata.X = np.clip(X, None, clip_norm_value)
    adata.uns['normalize_per_cell'] = True
adata