DL-designed proteins recruit coactivators with nanomolar affinity for precise transcriptional control
KevinKaichuang · x · 2026-10-23
Transcription factors recruit coactivators via weak interactions, making synthetic, precise transcriptional modulation hard. This work instead uses deep learning to design high-affinity coactivator-recruiting proteins.
Pipeline highlights:
- AlphaFold3 pairwise screening of 12 coactivator domains × 726 TF TAD regions (6,520 windows) to locate interaction interfaces
- RFdiffusion for backbone generation, ProteinMPNN for sequence design, AlphaFold3 for filtering
- Designed binders against p300-KIX, MED25, p300-TAZ2 and others with SPR-confirmed nanomolar Kd
- Pooled selection confirms the binders activate or repress gene expression with coactivator specificity in cells
A new AI-driven route to synthetic transcriptional control and gene therapy applications.
Related event: Baker Lab Designs De Novo Transcriptional Activators(3 posts)→
More from Research
- Modular low-cost UMI-style data collection rig inspired by fastumi and Kiwi — k7agar · 2026-10-23
- Sudo L7 benchmark uses 76-word natural prompts vs 291-427 words for rivals — echen · 2026-10-23
- 8B safety classifier trained on 1,275 prompts beats GPT-5.4 across 5 benchmarks — HaggaiMaron · 2026-10-23
- Sergey Levine on founding Physical Intelligence: robots grasping all day in a Google conference room — adityaag · 2026-10-23
- Joint belief-decision training gains are fragile without matched eval formats — JessicaHullman · 2026-10-23
- RL fine-tuning Qwen3-8B with four reward designs to target belief vs. decision — JessicaHullman · 2026-10-23