🏆 Congrats to Dr. Jian Tang @tangjianpku on winning the 2026 ACM Test of Time Award!
"LINE: Large-scale Information Network Embedding" (WWW 2015) helped shape the field. Still a classic 10 years on.
Kudos to co-authors Meng Qu, Mingzhe Wang, Ming Zhang, Jun Yan & Qiaozhu Mei 👏
Congratulations to the @abscibio team on the release of Origin-1!
Exciting to see the global AI–protein community pushing the boundaries of what’s possible together. We’re excited to be part of this moment!Looking forward to more progress ahead.🚀
We @abscibio are excited to share Origin-1, our platform for de novo antibody design! We select "zero-prior" epitopes without guidance from or the availability of solved protein-protein complex structures and, using fewer than 100 designs per target, achieve success against four targets: COL6A3, AZGP1, CHI3L2, and IL36RA.
We validate binding, developability, and function via multiple orthogonal assays and confirm atomic accuracy with Cryo-EM for two targets.
We share detailed methods, including on the models underlying Origin-1, and release our data as a resource for the community.
📄 Read the preprint: absci.com/wp-content/uploads…
📊 Access our data: github.com/AbSciBio/origin-1
279
What if antibody discovery didn’t start in a lab — but in silico?
With GeoFlow V3, we bring “epitope-in, antibody-out” design to reality.
Test ≤50 designs and get a 15.5% hit rate within weeks.
🔗 Full article: biogeom.com/2025/10/22/epito…
Beyond antibodies.
The same all-atom framework extends to all proteins — from peptide binders to novel enzymes.
Biotech/pharma working on challenging epitopes or hard targets — let’s design and validate together.
Contact: [email protected]
#AIxBio #ProteinDesign #AntibodyDesign
375
Under the hood: an atomic diffusion model that
- Unifies structure prediction and design
- Improves high-DockQ rate by 45% vs. GeoFlow V2
- Supports precise epitope conditioning and structural cues
- Discriminates binders from non-binders (via ipTM)
1
1
467
De novo hits, fast.
Across 8 de novo campaigns (5 validated targets including TSLP, IL-33, IL-13, CCR8, PD-1), we test ≤50 designs per epitope — in a single round of characterization and find binders in 3 weeks.
Average wet-lab hit rate: 15.5% (50x higher than methods in 2024)
Longer thinking, fewer shots.
GeoFlow V3 doesn’t just generate binders — it generates, evaluates, and performs in silico evolution (iterative, partial redesign) on promising candidates which significantly improves bind rates.
1
2
529
🚀 Introducing 𝗚𝗲𝗼𝗙𝗹𝗼𝘄 𝗩𝟯 — our all-atom protein foundation model for "epitope-to-antibody" design.
From idea → verified binders in <3 weeks.
Test ≤50 designs and get a 15.5% hit rate.
Free demo: prot.design
Technical report: prot.design/report
BioGeometry retweeted
🔥(1/5) Introducing GeoFlow V2: A unified atomic diffusion model for protein design
-Unifies structure prediction & de novo design w/ versatile constraint support
-SOTA results in Ab:Ag folding & epitope-specific Ab design
Try it and read our report at prot.design!
BioGeometry retweeted
This week's PrimeView is inspired by the Primer on computational #ProteinDesign out yesterday. Using current knowledge on the physics and chemistry of molecules, machine learning tools can be used to unlock doors to new worlds of proteins.
📖👉:go.nature.com/41zHPaw
Our GeoBiologics platform boosted the 8G3 antibody’s neutralizing activity by 1000-1500x against JN.1. The study is now out in
@PNASNews
, showing AI’s transformative role in antibody engineering. #AI4Science #drugdiscovery [Study]🔗pnas.org/doi/10.1073/pnas.24…
281
Replying to @talaldotpdb @richardwshuai
Impressed by the results achieved in just 5-10 steps! We definitely need to give more attention to sidechain structure—they play a crucial role in binding affinity.
102
Replying to @ScienceMagazine
Impressive cover! Programmable proteins could have a wide range of applications in the future.
27
Connect your biologics development to AI with GeoBiologics!
🔬 Harness cutting-edge AI for de novo protein design
🎯 Achieve multi-objective optimization w/ ease
🛠️ Customize solutions to fit your needs
✨ Enjoy an intuitive user experience
Try now at geobiologics.biogeom.com/
Excited to deepen our strategic cooperation with @Sino_Biological! Combining our protein foundation model GeoFlow with Sino Biological's protein expression and wet lab platform, we aim to set new standards in protein R&D. Learn more at biogeom.com/2024/09/03/bioge…
💡Introducing GeoFlow: cutting-edge protein generative model with:
1️⃣SOTA Ab-Ag docking success rates: ≈ AF3, 2x AF2
2️⃣Exceptional design capabilities: GeoFlow generated mAbs w/ novel HCDR3 that have up to 3x binding to the same epitope
🔗Try now at geobiologics-lite.biogeom.co…
📢After a $10M angel round, we're excited to announce our latest funding round led by Jiangmen Ventures. Funds will accelerate our work in GenAI for protein design w/ applications biomedicine and synthetic biology. 🚀
Stay tuned for more innovations! 🌟
biogeom.com/2024/07/22/bioge…
BioGeometry retweeted
Will be at #NeurIPS2023 next week. Happy to meet and chat together about #AIforDrugDiscovery, #GeometricDeepLearning, and #GenerativeModels.
Hello! Over the next week, we will be highlighting our invited speakers and authors.
Our first speaker is @tangjianpku @Mila_Quebec. Prof Tang works on generative AI and geometric deep learning for protein design. He will be opening the session at 8:30 AM. See you there!



















