Back

GeroEngine: Generative single-cell aging trajectories reveal a bidirectionally traversable identity core and direction-specific inflammatory remodeling

Bhak, Y.; Jeon, S.; Bhak, J.

2026-06-11 bioinformatics
10.64898/2026.06.08.731002 bioRxiv
Show abstract

Single-cell RNA sequencing (scRNA-seq) maps aging tissues at high resolution but is destructive, preventing longitudinal tracking; dropout and zero-inflation artifacts, amplified by shift-invariant linear simulations, confound age-associated variability. We developed Gero-Engine, a technical-artifact-aware framework combining VAE-based trajectory simulation, LOPO cross-validation, linear baselines, reverse traversal, and reverse-directed network inference. In microglia and HSCs, the VAE reduced technical-artifact carryover while preserving trajectory heterogeneity and improving alignment to artifact-reduced reference manifolds. Consensus GeroTargets and GeroRegulators defined tissue-specific GeroNetworks organized into three pillars: lineage/replication identity collapse, a sex-dimorphic endocrine/stress core, and inflammatory remodeling. Forward and reverse simulations aligned to the common young[->]old aging axis revealed a sign-coherent, direction-specific program: identity/replication targets were bidirectionally recovered, whereas MHC/NF-{kappa}B inflammatory programs were preferentially forward-recovered. These results support identity collapse as a deep traversable core of aging and nominate upstream homeostatic restoration over downstream inflammatory suppression.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 15%
11.9%
Similar papers in this journal
2
Nature Aging
○ 60 papers in training set
Top 0.1%
11.9%
Similar papers in this journal
3
PLOS Computational Biology
● 1863 papers in training set
Top 6%
6.3%
Similar papers in this journal
4
Cell Systems
○ 201 papers in training set
Top 0.7%
5.5%
Similar papers in this journal
5
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 10%
4.9%
Similar papers in this journal
6
eLife
● 5828 papers in training set
Top 30%
4.0%
7
Cell Reports
○ 1498 papers in training set
Top 10%
4.0%
8
GeroScience
○ 109 papers in training set
Top 0.6%
4.0%
50% of probability mass above
9
Advanced Science
○ 286 papers in training set
Top 2%
3.2%
10
Aging Cell
○ 165 papers in training set
Top 0.9%
3.2%
11
npj Aging
○ 22 papers in training set
Top 0.1%
2.4%
12
Nature Medicine
○ 125 papers in training set
Top 1%
2.1%
13
Briefings in Bioinformatics
◐ 354 papers in training set
Top 4%
2.1%
14
Science Advances
● 1243 papers in training set
Top 17%
2.1%
15
Nature Methods
○ 385 papers in training set
Top 4%
1.7%
16
Autophagy
○ 39 papers in training set
Top 0.3%
1.5%
17
Nucleic Acids Research
◐ 1281 papers in training set
Top 10%
1.4%
18
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 4%
1.3%
19
Scientific Reports
○ 3612 papers in training set
Top 60%
1.3%
20
Genome Biology
○ 637 papers in training set
Top 7%
1.3%
21
Science
● 477 papers in training set
Top 7%
1.1%
22
iScience
○ 1154 papers in training set
Top 26%
1.1%
23
Genome Medicine
○ 183 papers in training set
Top 4%
1.1%
24
Molecular Systems Biology
○ 162 papers in training set
Top 2%
1.1%
25
Nature Metabolism
○ 69 papers in training set
Top 2%
1.0%
26
Aging
○ 75 papers in training set
Top 2%
0.8%
27
Communications Biology
○ 993 papers in training set
Top 30%
0.8%
28
Genome Research
● 468 papers in training set
Top 7%
0.6%
29
Patterns
○ 78 papers in training set
Top 3%
0.6%

"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.