Multiscale learning of gene network-driven phenotypic dynamics of single cells
Dongyan Zhang, Jinan Li, Qing Nie, Xiaoqiang Sun · 2026
Read Multiscale learning of gene network-driven phenotypic dynamics of single cells by Dongyan Zhang, Jinan Li, Qing Nie, Xiaoqiang Sun (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13434742.txt.
Description
Zhang, Li, Nie, and Sun present GRNvelo, a framework that rebuilds how gene regulatory networks steer cells toward their fates over time. It uses time-series single-cell RNA sequencing. The model couples network-driven velocity for single cells with nonlocal population dynamics. Those dynamics track cell numbers growing and shrinking across a developing tissue.
Two physics-guided neural networks pair up. One jointly infers network velocity and each cell's latent time. The second refines velocity against the population-level picture of growth. Benchmarks covered two fake data sets and four real ones. They included branching development and varied responses to gene tweaks. GRNvelo beat tools in the reported tests. Beyond rebuilding cell paths, the framework predicted how cell fates shift under genetic tweaks. Those included combined changes never present in the training data. The authors frame it as an approach for predicting and steering cell fate.
Catalog record
| CREATOR | Dongyan Zhang, Jinan Li, Qing Nie, Xiaoqiang Sun |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 218 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |