Deep learning predicts age-related chromatin states in mouse blood stem cells
ChromAgeNet, a deep learning model trained on 3D microscope images of mouse blood stem cell nuclei, distinguished young from aged cells with an AUROC of 0.77 ± 0.03 and flagged chromatin features as predictive markers. [1]
Automatic summary · not editorially reviewed
The key findings
ChromAgeNet was trained on 3D microscope images of DAPI-stained mouse hematopoietic stem cell nuclei to distinguish young from aged cells. [1]
The model achieved an AUROC of 0.77 ± 0.03 and outperformed classical machine learning models using handcrafted chromatin features from the same dataset. [1]
Using explainable artificial intelligence, the authors identified chromatin entropy, peripheral heterochromatin and chromatin condensates as predictive markers. [1]
The source at a glance
- Study type / format
- Abstract of a preclinical computational and imaging study
- Population / context
- Murine hematopoietic stem cells, including aged HSCs treated with epigenetic drugs in a proof-of-concept screen
- Summary based on
- Study abstract
- Original publication
- October 1, 2026
Mouse stem cell nuclei imaged to model aging chromatin
The authors note that chromatin architecture changes are a well-established hallmark of aging but remain largely unexplored as quantitative markers. They built ChromAgeNet, an interpretable deep learning approach using convolutional neural networks, to learn changes in spatial features of chromatin architecture as hematopoietic stem cells (HSCs) age. The model was trained on 3D microscope images of DAPI-stained HSC nuclei to tell young from aged murine HSCs apart.
Supporting source [1]
AUROC 0.77 and three flagged chromatin markers
The model reached an AUROC of 0.77 ± 0.03 and outperformed classical machine learning models trained on handcrafted chromatin features from the same dataset. Using explainable artificial intelligence techniques, the authors identified chromatin entropy, peripheral heterochromatin and chromatin condensates as predictive markers. As a proof of concept, they also evaluated the model as a phenotypic screening tool for aged HSCs treated with epigenetic drugs to detect rejuvenation.
Supporting source [1]
Abstract-only scope: mouse cells, no human lifespan data
The abstract reports a model that predicts age-associated chromatin states in the hematopoietic compartment of mice. The drug-screening use is described as a proof of concept, not a validated therapy or screening product. The abstract does not report sample sizes, how many animals were used, or any outcome in humans such as lifespan or healthspan. Only the abstract was reviewed; the full text was not reviewed.
Supporting source [1]
How to interpret this
Abstract only; the full text was not reviewed. The work concerns murine HSCs and a proof-of-concept drug-screening evaluation; no human outcomes, sample sizes or follow-up are reported in the abstract.
Based on the study abstract; the full paper, tables and supplementary data were not reviewed.
Source & supporting passages
Picazo PI, Mejía-Ramírez E, Di Bari D, Vitali E, Florian MC, Petrone P. · europepmc.org · October 1, 2026