🧬 Fusion of protein sequence and structural information, using denoising pre-training network for zero-shot protein engineering (eLife 2025).

Directed evolution of proteins in sequence space with gradients

Protein Design by Machine Learning guided Directed Evolution

LatentDE: Latent-based Directed Evolution for Protein Sequence Design

TeBaAb: Text-Based Antigen-Conditioned Antibody Redesign via Directed Evolution

Computational model of laboratory directed evolution + experiments.

Learning to Solve Multiresolution Matrix Factorization by Manifold Optimization and Evolutionary Metaheuristics

Directed Evolution of Proteins via Bayesian Optimization in Embedding Space

Unified Python library and CLI for protein structure prediction and inverse folding.

Directed Evolution in Silico

GenLoop — a closed generative drug-design loop on dtSFM; directed evolution of small molecules (Reddy 2026). Generate → encoder-rerank → AlphaFold-3 verify → LoRA-refine.

ProtEvolveLab — Interactive Protein Language Model & In-Silico Directed Evolution Laboratory. 6 modules: amino acid embeddings, NK fitness landscapes, directed evolution simulator, epistasis mapper, MCTS protein designer, strategy arena. Zero dependencies.

Generates randomly generated fastQ files from a template and upstream sequence.

GFlowNets for in-silico directed evolution: sampling diverse, feasible, high-fitness sequence variants rather than hill-climbing to one. Exact-ground-truth benchmarks (Ehrlich functions, GB1), multi-objective rewards, and a design-build-test-learn campaign loop under an evaluation budget.

Calculates the probability of finding the top variant in a library of sequences.

A protocol for directed code evolution orchestration, with a reference implementation and conformance suite