Back

In vivo measurement and prediction of cell-type-resolved non-coding variant effects in human immunity

Liu, Q.; Zhang, J.; Zhu, Y.; Wang, K.; Zhang, C.; Zhang, J.; Jin, Y.; Zhang, Y.; Liu, Z.; Guo, Y.; Yang, J.; Liu, Z.

2026-05-06 genetic and genomic medicine
10.64898/2026.05.05.26352448 medRxiv
Show abstract

Interpreting non-coding variant effects in their native cellular context is the central problem in medical genomics. Using paired single-cell ATAC-seq and whole-genome sequencing in human immune cells, we measured allele-specific regulatory effects at 129,241 heterozygous variants across seven cell types. We trained REG-FOCUS, a transferable sequence-and-chromatin framework, and extended predictions to 1,188,417 common variants across twelve immune cell types. These predictions revealed that enhancer-centric variant directionality is governed by quantitative tuning of transcription factor binding affinity. Applied to systemic lupus erythematosus, REG-FOCUS prioritized causal regulatory variants selectively converging on pathways of approved therapies, and enabled us to identify rare enhancer variants in patients corroborated by expression outliers in carriers. This work brings cell-type-resolved variant interpretation to non-coding personal genomes.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

1
Science
● 477 papers in training set
Top 0.2%
18.6%
Similar papers in this journal
2
Cell Genomics
○ 172 papers in training set
Top 0.1%
13.0%
Similar papers in this journal
3
Nature Genetics
○ 286 papers in training set
Top 0.4%
12.8%
Similar papers in this journal
4
Nature Communications
○ 5641 papers in training set
Top 15%
12.0%
Similar papers in this journal
50% of probability mass above
5
Nature
○ 645 papers in training set
Top 3%
5.5%
Similar papers in this journal
6
Cell
○ 431 papers in training set
Top 2%
4.1%
7
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 16%
3.2%
8
Science Advances
● 1243 papers in training set
Top 10%
3.2%
9
Nature Immunology
○ 79 papers in training set
Top 1%
2.4%
10
Nature Neuroscience
○ 252 papers in training set
Top 3%
2.4%
11
eLife
● 5828 papers in training set
Top 45%
2.1%
12
Genome Medicine
○ 183 papers in training set
Top 3%
1.7%
13
Genome Biology
○ 637 papers in training set
Top 6%
1.7%
14
Cancer Discovery
● 66 papers in training set
Top 1%
1.7%
15
Developmental Cell
○ 196 papers in training set
Top 3%
1.4%
16
Journal of Clinical Investigation
● 179 papers in training set
Top 4%
1.3%
17
Nature Structural & Molecular Biology
236 papers in training set
Top 3%
1.1%
18
Nature Biomedical Engineering
○ 47 papers in training set
Top 2%
0.9%
19
Cell Reports
○ 1498 papers in training set
Top 26%
0.8%
20
Human Genetics and Genomics Advances
○ 84 papers in training set
Top 2%
0.8%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.