ProteinGym

ProteinGym benchmarks overview ProteinGym is an extensive set of Deep Mutational Scanning (DMS) assays curated to enable thorough comparisons of various mutation effect predictors indifferent regimes. It is comprised of two benchmarks: 1) a substitution benchmark which consists of the experimental characterisation of ∼

Papers using ProteinGym (17)

Can Tabular In-Context Learners Generalize to Biomolecular Property Prediction?2026Metalic: Meta-Learning In-Context with Protein Language Models2024 · 3 citesA Diffusion Model to Shrink Proteins While Maintaining Their Function2025 · 2 citesAugmenting Biological Fitness Prediction Benchmarks with Landscapes Features from GraphFLA2025 · 1 citesTriFit: Trimodal Fusion with Protein Dynamics for Mutation Fitness Prediction2026Reverse Distillation: Consistently Scaling Protein Language Model Representations2026Binary Latent Protein Fitness Landscapes for Quantum Annealing Optimization2026Rank-and-Reason: Multi-Agent Collaboration Accelerates Zero-Shot Protein Mutation Prediction2026No Generation without Representation: Efficient Causal Protein Language Models Enable Zero-Shot Fitness Estimation2026Evolutionary Profiles for Protein Fitness Prediction2025A Matrix Variational Auto-Encoder for Variant Effect Prediction in Pharmacogenes2025Protein Language Model Zero-Shot Fitness Predictions are Improved by Inference-only Dropout2025Exploring zero-shot structure-based protein fitness prediction2025Exploring Large Protein Language Models in Constrained Evaluation Scenarios within the FLIP Benchmark2025Multi-Scale Representation Learning for Protein Fitness Prediction2024 · 14 citesFine-tuning Protein Language Models with Deep Mutational Scanning improves Variant Effect Prediction2024 · 11 citesRetrieval-Enhanced Mutation Mastery: Augmenting Zero-Shot Prediction of Protein Language Model2024 · 4 cites