De novo design of light-regulated dynamic proteins using deep learning
BARTH, P.; Scutteri, L.; Abriata, L.; Zhang, S.; Hacisuleyman, A.; Lau, K.; Pojer, F.; Rahi, S. J.
Show abstract
Recent advances in deep learning have enabled accurate design of static protein structures, but the de novo design of protein functions controlled by programmable, intramolecular conformational changes remains an unsolved challenge. Here, we present a general deep learning-guided framework for designing dynamic, multi-domain proteins allosterically regulated by light. By integrating photoresponsive domains into de novo scaffolds, we engineered conformational switches that exhibit precise, reversible structural transitions upon illumination. Structural, spectroscopic, and functional analyses validated our designs and demonstrated precise spatio-temporal optogenetic control of diverse cellular processes, including subcellular localization, intercellular signaling, and population-level behaviors. This work establishes a broadly applicable strategy for encoding long-range allosteric control through designed intramolecular motions, and opens new avenues for programming dynamic protein functions from first principles, with implications for basic research, synthetic biology, and therapeutic development.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Broadly applicable and accurate protein design by integrating structure prediction networks and diffusion generative models 97%
- Multiplexed, scalable analog recording of gene regulation dynamics over weeks using intracellular protein tapes 96%
- Atomically accurate de novo design of antibodies with RFdiffusion 95%
"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.