Research map / Label-efficient and robust learning / Recognition and long-tail
FLIP2: Expanding Protein Fitness Landscape Benchmarks for Real-World Machine Learning Applications
Kieran Didi, Sarah Alamdari, Alex Xijie Lu, Bruce James Wittmann, Kadina E Johnston, Ava P Amini, Ali Madani, Maya Czeneszew, Christian Dallago, Kevin K Yang
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 pretrained weights before fine-tuning (ablation) | CARP-640M · FLIP2 (16 splits) | Spearman: 13 of 16 splits improved | Improved |
| use pretrained weights before fine-tuning (ablation) | ESMC-300M · FLIP2 (16 splits) | Spearman: 10 of 16 splits improved | Mixed |
| supervised fine-tuning of zero-shot CARP-640M | CARP-640M · FLIP2 (16 splits) | Spearman: improved on 7/16 splits | Mixed |
| supervised fine-tuning on Amylase and by-position splits | CARP-640M · Amylase and by-position splits | Spearman | Hurt |
| add structure tokens (SaProt) to zero-shot scoring | SaProt-650M · FLIP2 zero-shot | Spearman | No clear effect |
| add structure tokens or LoRA to fine-tuning (ablation) | pLM fine-tuning · FLIP2 (16 splits) | Spearman | No clear effect |
| add zero-shot pLM scores as input to ridge regression (ablation) | ridge regression · Amylase, IRED, NucB | Spearman | Improved |
| add zero-shot pLM scores as input to ridge regression (ablation) | ridge regression · Hydro, Rhomax, PDZ3 | Spearman | No clear effect |
| add zero-shot scores to ridge regression on wild-type and position splits (ablation) | ridge regression · position and wild-type splits | Spearman | Improved |
| use FLIP2 splits instead of random splits | ridge regression · FLIP2 splits vs random splits | Spearman | Hurt |
| fine-tuned pLM vs simpler baselines | fine-tuned pLM · FLIP2 (16 splits) | Spearman | Hurt |
Closest papers by meaning
- Metalic: Meta-Learning In-Context with Protein Language Models (ICLR 2025 · similarity 0.88)
- Steering Generative Models with Experimental Data for Protein Fitness Optimization (NeurIPS 2025 · similarity 0.87)
- Learning to engineer protein flexibility (ICLR 2025 · similarity 0.87)
- DePLM: Denoising Protein Language Models for Property Optimization (NeurIPS 2024 · similarity 0.87)
- Improving protein optimization with smoothed fitness landscapes (ICLR 2024 · similarity 0.87)
- Augmenting Biological Fitness Prediction Benchmarks with Landscapes Features from GraphFLA (NeurIPS 2025 · similarity 0.87)
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.
