A Designed Carbamate Hydrolase
Our swing is a de-novo carbamate hydrolase with no prior art; that GPU run is deferred. What we designed in-silico is the LigandMPNN-redesign fallback: a known serine-hydrolase scaffold with its Ser-His-Asp triad repositioned for carbamate hydrolysis at near-neutral pH. An in-silico design, honestly a redesign, not a de-novo fold.
The CALB / PDB 1TCA serine-hydrolase scaffold our lead design (enzyme-003) is a LigandMPNN redesign of; its Ser-His-Asp catalytic triad (Ser105 / His224 / Asp187) is highlighted in green. This reference scaffold is downloadable in the readout. Drag to rotate.
Ser-His-Asp
engineered, near-neutral pH 7.0 to 8.5, cotton-compatible
In-silico PLACER preorganization targets for the LigandMPNN redesign, on the Lauko thresholds. Design metrics, not an expressed protein.
redesign-not-de-novo: LigandMPNN redesign of Candida antarctica lipase B scaffold (1TCA) leveraging alpha/beta-hydrolase fold for polycarbonate-diol carbamate cleavage. TM 0.85 to template.
best TM-score 0.85
H12MDI/polycarbonate-diol/BDO fiber paired with CalB-derived lipase. The polycarbonate soft segment provides hydrolytic stability in use but cleaves enzymatically under controlled conditions, recovering carbonate-diol monomers.
PUase-H12MDI-v1 → polycarbonate carbamate linkage
One inverse-design problem, two coupled molecules. The amber fiber's carbamate is the exact bond the green enzyme is built to cut: the substrate detaches, docks into the Ser-His-Asp pocket, the bond cleaves to recoverable monomers, and the two accents fuse at the cut. Proof it is one loop, not two projects.