Answers are grounded in this paper and cite their sources.
Experimental study reveals zero-shot protein language models identify functional mutations in ChrimsonR, suggesting computational models can efficiently optimize optogenetic tools.
A central challenge in opsin engineering is identifying mutations that reliably improve desired functional properties, a task made difficult by the enormous mutation space and limited throughput of electrophysiological screening. Improving opsin properties such as photocurrent amplitude and light sensitivity have the potential to broaden the use of opsins to low-light and deep-tissue applications. With this goal, we applied zero-shot protein language models (ESM-1b/1v) to recommend ChrimsonR mutations and experimentally validated all 17 of these variants using whole-cell patch clamp electrophysiology (n=6 cells per mutation). Despite many mutations reducing function, protein language models identified both known functional residues and unconventional substitutions that produced large functional gains and synergized with K176R to improve kinetics. Two mutations, E300G and E300P, increased sustained photocurrents from 66 pA (control) to 305 pA and 255 pA at 635 nm, reduced EC 50 at 575 nm from 0.19 mW to 0.07 mW, and altered kinetics ( τ off increased from 0.06 s up to 0.40 s). Our results suggest that protein language models, even without task-specific training, can be used alongside electrophysiological measurements as a strategy for screening opsins for enhanced photocurrent.
No takes yet. Share an insight, caveat, or question.
Ehrlich et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: