Activity-Oriented Enzyme Sequence Design by Steered Inverse Protein Folding
Code, data and notebooks for engineering enzymes with ultra-high throughput microfluidics. Please read our accompanying paper for details.
RFdiffusion -> ProteinMPNN -> ESMFold pipeline generating candidate enzymes; 30 shortlisted designs against fluoroacetate dehalogenase's active site as a stand-in PFAS target
A robust and scalable framework for resource-constraint models through stochastic tuning of deep learning-predicted kinetic parameters
Protein language model embeddings, rank-based prediction, and active learning for prioritizing Rubisco variants with desired traits.
Open-source Sci-Evo dataset release for AGI4S: protein/enzyme engineering trajectories, code, source papers, and MinerU artifacts.
Code and minimal reproducibility data for RIAD/RIDD productive spatial coupling and PSCI-PairRank.
Reproducible data and ML workflows for enzyme mutation analysis.
Educational prototype: AI-assisted enzyme candidate prioritization for PET plastic degradation. Synthetic data, transparent scoring, real AlphaFold/UniProt structural anchors. Not a validated scientific system.
Three-modal enzyme function classification using sequence, structure, and text representations.
ProteinGym benchmark harness and ESM-2 zero-shot mutational effect prediction engine for biocatalysis.
This project combines in silico engineering of PETase/MHETase with NGS-based metagenomics to improve plastic biodegradation. Using molecular docking, mutation design, and sequence analysis, it identifies enhanced enzyme variants and novel PETase-like candidates from environmental samples
Bayesian optimization for enzyme reaction conditions — Gaussian process surrogate, expected improvement acquisition, Chromeleon HPLC integration, and iterative experiment design
Reproducible pipeline that turns an enzyme structure into a ranked, screenable set of mutation targets for improving catalytic activity. Validated against 100 published β-glucosidase mutants, then applied to an uncharacterised fungal arabinofuranosidase.
Cradle — independent third-party profile of a public API surface, by API Evangelist. Cradle is a Dutch–Swiss biotechnology software company founded in 2021 with headquarters in Amsterdam and an office in Zurich, building an AI-powered protein engineering platform for industrial biotech and biopharma R&D teams.
To associate your repository with the enzyme-engineering topic, visit your repo's landing page and select "manage topics."