Kyle Tretina, Ph.D.

🧬 DNA-binder design could make gene targeting less constrained by natural protein scaffolds

📊 High-thinking RFD4-Proteina had a 3.5% computational clustered pass rate versus 0.8% for fixed-DNA BoltzGen across three targets

If you are designing sequence-specific gene-control tools, you should check this out

⚗️ Enzyme design is expanding the chemistry we can perform 🧪

For enzyme scaffolding, high-thinking RFD4-Proteina achieved a 12.1% computational clustered pass rate versus 5.7% for AlphaProtein-novo codesign 📊

Please test this model on custom industrial catalysts and share your results here!

🧪 Designed binders could expand chemical sensing beyond natural receptors

High-thinking RFD4-Proteina generated at least one design passing computational structural checks for each of 77 benchmark targets

Can't wait to see the custom biosensors people design for drugs and metabolites with this! 🔬

🧩 Protein design is moving from folds to molecular recognition

High-thinking RFD4-Proteina achieved 42x higher generation throughput than the benchmark's single-backbone BindCraft, with a higher computational clustered pass rate ⚡

Protein binder design just took a jump forward

Is RFD4-Proteina the biggest AI x Bio news to come out today?