arXiv:2605.28111v1 Announce Type: new Abstract: Predicting how a cell will change its transcriptional state under a developmental signal or a genetic perturbation is the computational core of in-silico biology and the AI Virtual Cell program. Existing approaches either fit static control-to-treated maps that discard time, or solve multi-step ODE / Schr\"odinger-bridge problems on each dataset independently. We introduce Chreode, a one-step cell world model that predicts action-conditioned cell-state transitions through a structured residual transition operator. It shifts distributional evoluti
The accelerating pace of AI development allows for more sophisticated computational models that can tackle complex biological systems, moving beyond static analyses.
This development is crucial for advancing in-silico biology and the AI Virtual Cell program, enabling more accurate and dynamic predictions of cellular behavior under various conditions.
Traditional static or multi-step models for cellular state prediction are being superseded by more integrated, one-step approaches that can better simulate temporal dynamics and perturbation responses.
- · Biopharmaceutical companies
- · Synthetic biology researchers
- · AI Virtual Cell program initiatives
- · Drug discovery methods relying on slower, less predictive models
- · Biological research limited by static simulations
Chreode enables faster and more accurate computational modeling of cellular responses to genetic alterations and developmental signals.
This could significantly shorten drug discovery timelines and improve the design of engineered biological systems.
Ultimately, it may lead to new therapeutic strategies and the creation of entirely novel biological functions and organisms.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG