Research map / Image generation / Autoregressive and latent generation
Towards A Generative Protein Evolution Machine with DPLM-Evo
Xinyou Wang, Liang Hong, Jiasheng Ye, Zaixiang Zheng, Shujian Huang, Quanquan Gu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use contextualized evolutionary noising kernel instead of uniform corruption (ablation) | DPLM-Evo · ProteinGym substitution (avg Spearman) | avg Spearman: 0.295 -> 0.42 | Improved |
| use contextualized kernel instead of BLOSUM kernel (ablation) | DPLM-Evo · ProteinGym substitution (avg Spearman) | avg Spearman: 0.35 -> 0.42 | Improved |
| use BLOSUM kernel instead of uniform kernel (ablation) | DPLM-Evo · ProteinGym substitution (avg Spearman) | avg Spearman: 0.295 -> 0.35 | Improved |
| enable indel sampling in GFP directed evolution (ablation) | DPLM-Evo · GFP directed evolution (20 iterations) | pTM: 0.793 -> 0.857 | Improved |
| use DPLM-Evo instead of ESM-2 as proposer under same beam search | GFP directed evolution · GFP directed evolution (20 iterations) | pTM: 0.737 -> 0.793 | Improved |
| scale DPLM-Evo from 650M to 3B | DPLM-Evo · ProteinGym substitution | Spearman | Improved |
| scale ESM-2 from 650M to 3B | ESM-2 · ProteinGym substitution | Spearman: approximately 0.01 lower | Hurt |
Closest papers by meaning
- DePLM: Denoising Protein Language Models for Property Optimization (NeurIPS 2024 · similarity 0.92)
- Diffusion Language Models Are Versatile Protein Learners (ICML 2024 · similarity 0.91)
- DPLM-2: A Multimodal Diffusion Protein Language Model (ICLR 2025 · similarity 0.90)
- MutaPLM: Protein Language Modeling for Mutation Explanation and Engineering (NeurIPS 2024 · similarity 0.88)
- Steering Protein Language Models (ICML 2025 · similarity 0.88)
- LEMON-ZEST: Evolution-Informed Tokenization for Efficient Protein Language Modeling (NeurIPS 2026 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
