---
library_name: pyaging
tags:
- pyaging
- aging-clock
- biology
- dna-methylation
---

# celldrift

Replication signature trained on 31 serially passaged hTERT-immortalized human astrocyte samples. The released 62-PC elastic-net predictor over 2,322 CpGs is folded algebraically into a single linear score. Output is calibrated to cumulative population doublings, not years. Original-author missingness is preserved: supplied NaNs become zero, whereas absent CpGs use module means. GitHub commit 066b3e816795f3c0b372ff8d33f59185a490bd8b; DOI 10.5281/zenodo.7693699. Upstream software license is unspecified (Zenodo: Other (Open)).

Model weights retain the original authors' terms; the pyaging software license does not relicense them.

| | |
|---|---|
| **Predicts** | mitotic age |
| **Species** | Homo sapiens |
| **Tissue** | cultured human cells |
| **Data type** | DNA methylation |
| **Model type** | PCA + elastic net regression |
| **Year** | 2023 |

## Use with pyaging

```python
import pyaging as pya

pya.pred.predict_age(adata, ["celldrift"])
```

Browse every clock in the [pyaging Clock Catalogue](https://pyaging.readthedocs.io).

## Citation

Minteer, C. J., et al. (2023). More than bad luck: Cancer and aging are linked to replication-driven changes to the epigenome. Science Advances, 9(29), eadf4163.

https://doi.org/10.1126/sciadv.adf4163