iiinova — where your digital self keeps evolving

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Find all details, links to papers, model weights and code, and link to the Teddymer dataset on our project page. 🔥 Project page: research.nvidia.com/labs/genair/... Finally, please check Kieran's (@kdidi.bsky.social) great thread on Proteina-Complexa, too: bsky.app/profile/kdid...

research.nvidia.com

Proteina-Complexa

Proteina-Complexa combines joint sequence-structure generation in a continuous latent space with inference-time search for de novo protein binder design. In a massive-scale experimental benchmark — o...

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Grateful to our amazing academic and industry partners for this exciting collaboration! @manifoldbio.bsky.social, @novonordisk.bsky.social, Viva Biotech, Duke University (Soderling Lab), Cambridge University (Hollfelder Lab), @lmu.de (Khmelinskaia Group), @SeoulNatlUni (Steinegger Lab) (19/n)

110

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸This was an effort by a brilliant team at NVIDIA & partners NVIDIA shoutouts: bsky.app/profile/kdid..., Danny Reidenbach, Zuobai Zhang, Guoqing Zhou, Zhonglin Cao, Tomas Geffner, Micha Livne, @machine.learning.bio, Emine Kucukbenli, @arashv.bsky.social A privilege working with this team. (18/n)

bsky.app

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 A key highlight: We achieved de novo design of carbohydrate binders. We targeted the Blood Group B antigen, reaching a 21% hit rate. To the best of our knowledge, no prior computational methods have achieved de novo design against these challenging polar targets. (17/n)

110

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Targeting Viruses: As part of the recent Adaptyv binder competition, we used Proteina-Complexa to design a nanomolar binder (56 nM) against the Nipah virus G protein, successfully targeting its recessed receptor-binding site (see visualization of the experimental hit). (16/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Kinase targets and varying binder size: Spanning two distinct size regimes, we designed peptide (<31 amino acids) and miniprotein binders (49–74 amino acids) for kinase targets like CK1δ and PAK1. Proteina-Complexa achieved 40-50% hit rates for these difficult motifs. (15/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Activin Binders: Next, we designed de novo binders for the Activin receptor type IIA (ActRIIA) that block myostatin signaling in cells. Our tightest binder showed a KD of 36 nM and functional inhibition in downstream experiments. (14/n)

110

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Aside from the massive screen, we conducted case studies of different targets in separate experiments. For instance, we generated de novo binders for PDGFR: We achieved a very high 63.5% hit rate, with top candidates reaching double-digit picomolar affinity (93.6 pM). (13/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 As part of this large screen, we also conducted a large-scale systematic wet lab comparison to contemporary binder design methods. Proteina-Complexa outperforms the baselines; see chart below. In particular its self-generated sequences work well - no more re-design. (12/n)

110

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 First, we performed a high-throughput massive-scale screen across 127 diverse and challenging targets. We screened around 1 million candidates in total, measuring all-to-all binding. 86 targets yielded experimentally validated hits, with 74 of those being specific. (11/n)

110

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Now experimental results. We conducted a massive campaign in collaboration with @ManifoldBio, @viva_biotech, @novonordisk, @Cambridge_Uni, @DukeU, @LMU_Muenchen. 🧬 "Latent Generative Search unlocks de novo Design of Untapped Biomolecular Interactions at Scale." (10/n)

110

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Quantitatively, our inference-time scaling strategies (we use, for instance, MCTS, Feynman-Kac Steering and Beam Search) outperform previous hallucination methods under normalized compute budgets, setting a new state-of-the-art in in-silico binder design. (9/n)

110

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Beyond regular binding, our model excels at atomistic motif scaffolding for enzyme design. On the AME benchmark, Proteina-Complexa significantly outperforms RFDiffusion2, faithfully reconstructing complex active site geometries. (8/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 We can explicitly optimize for biophysical properties during generation. In particular, Proteina-Complexa leverages interface hydrogen bond optimization, which steers the generative search toward candidates with denser, more stable interaction networks. (7/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 To overcome the scarcity of multimer data, we introduce Teddymer: a synthetic dataset of 0.5M clustered binder-target pairs, constructed by splitting AFDB monomers into structural domains, simulating realistic protein-protein interactions. Link to data on project page. (6/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 "Latent Generative Search": We unify generative modeling with test-time optimization for high performance protein design. Scaling compute at inference via strategies like beam search and MCTS, we can steer the model toward higher-quality, physically realistic binders. (5/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Technical Core: We use La-Proteina's partially latent flow matching framework that co-designs protein sequence and atomistic structure jointly. No discrete amino acid tokens. No separate inverse folding (no sequence re-design). Fully end-to-end atomistic generation. (4/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 Proteina-Complexa is a new protein binder design framework leveraging generative pretraining and test-time compute scaling. It can generate in silico binder candidates for diverse targets, including single and multi-chain proteins and small molecule ligand targets. (3/n)

101

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

🔸 We present two papers covering Complexa's core method development and a large-scale experimental validation effort. Let’s dive into the ICLR 2026 Oral paper on the method first: "Scaling Atomistic Protein Binder Design with Generative Pretraining and Test-Time Compute" (2/n)

100

Karsten Kreis @karstenkreis.bsky.social · 17/03/2026

📢📢 Proteina-Complexa 📢📢 Atomistic Binder Design with Generative Pretraining and Test-Time Compute + Experimental Validation at Scale ⭐️ Project page (research.nvidia.com/labs/genair/...) for: 📜 Method paper (ICLR' 2026 Oral) 🧬 Wet lab paper 🛠️ Code & Models 📁 Data 🧵 Thread (1/n)

1112

Reposted by Karsten Kreis

Kevin K. Yang 楊凱筌 @kevinkaichuang.bsky.social · 16/07/2025

Partially-latent flow matching enables sequence-structure codesign of large proteins and functional motif scaffolding. @kdidi.bsky.social @machine.learning.bio @karstenkreis.bsky.social @arashv.bsky.social arxiv.org/html/2507.09...

1258

Karsten Kreis @karstenkreis.bsky.social · 24/04/2025

3⃣ Efficient Molecular Conformer Generation with SO(3) Averaged Flow-Matching and Reflow openreview.net/forum?id=1B1... ⭐️ I'm also on a panel on synthetic data (synthetic-data-iclr.github.io)! I'm excited to discuss research and to meet new and old friends and collaborators! 🎉 (5/n)

000

Karsten Kreis @karstenkreis.bsky.social · 24/04/2025

@gembioworkshop.bsky.social papers: 1⃣ EquiJump: Protein Dynamics Simulation via SO(3)-Equivariant Stochastic Interpolants arxiv.org/abs/2410.09667 (oral) (screenshot below) 2⃣ Hierarchical Protein Backbone Generation with Latent and Structure Diffusion arxiv.org/abs/2504.09374 (4/n)

100

Karsten Kreis @karstenkreis.bsky.social · 24/04/2025

3⃣ Energy-Based Diffusion Language Models for Text Generation arxiv.org/abs/2410.21357 Posters 2 4⃣ Truncated Consistency Models arxiv.org/abs/2410.14895 Posters 4 (screenshot below) (3/n)

100

Karsten Kreis @karstenkreis.bsky.social · 24/04/2025

Main track: 1⃣ Proteina: Scaling Flow-based Protein Structure Generative Models research.nvidia.com/labs/genair/... Orals 3B, posters 4 (video below) 2⃣ ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids arxiv.org/abs/2503.05025 Orals 2C, posters 3 (2/n)

100

Karsten Kreis @karstenkreis.bsky.social · 24/04/2025

🔥 I'm at ICLR'25 in Singapore this week - happy to chat! 📜 With wonderful co-authors, I'm co-presenting 4 main conference papers and 3 @gembioworkshop.bsky.social papers (gembio.ai), and I contribute to a panel (synthetic-data-iclr.github.io). 🧵 Overview in thread. (1/n)

131

Karsten Kreis @karstenkreis.bsky.social · 11/03/2025

🔥 ProtComposer (ICLR'25 Oral) is a Swiss Army knife: (i) Manually create new protein structure layouts? ✅ (ii) Generation with favorable designability/diversity/novelty trade-offs? ✅ (iii) Spatially edit given proteins? ✅ Very original work by the amazing @hannes-stark.bsky.social and Bowen Jing!🔥

072

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸Check out our project page (research.nvidia.com/labs/genair/...), our paper (arxiv.org/abs/2503.00710), and our code (github.com/NVIDIA-Digit...). 🔥 We released 8 sets of weights, for all experiments, for you to play with! 🔥 Enjoy! And see you at ICLR'25! 😀 (11/11)

research.nvidia.com

Proteina: Scaling Flow-based Protein Structure Generative Models

020

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸Proteina is a fantastic collaboration with wonderful colleagues at NVIDIA: 🔥 Tomas Geffner*, @kdidi.bsky.social*, Zuobai Zhang*, Danny Reidenbach, Zhonglin Cao, @jyim.bsky.social , Mario Geiger, @machine.learning.bio, Emine Kucukbenli, @arashv.bsky.social, @karstenkreis.bsky.social* 🔥 (10/n)

171

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸We also demonstrate LoRA-based fine-tuning on a smaller set of high-quality protein structures from the PDB, and we show that autoguidance, where the model is guided by a weaker version of itself, can be used to boost designability. See our paper for details. (9/n)

120

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸Protein structure generation performance is often measured in terms of designability, diversity and novelty. Drawing inspiration from image generation, we explore three complementary metrics that analyze models at the distribution level, providing additional insights. (8/n)

120

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸Proteina also outperforms previous models on motif-scaffolding, where a functionally relevant motif is given and the model is tasked with generating a viable supporting scaffold. Below, we show quantitative evaluations for the benchmark introduced by RFDiffusion. (7/n)

110

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸Quantitatively, Proteina achieves state-of-the-art designable and diverse protein backbone generation (unconditional or fold class-conditional). In particular at long lengths, it significantly outperforms previous models, which cannot generate proteins at this scale. (6/n)

110

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸Proteina uses an efficient and scalable non-equivariant transformer network with up to 400M parameters. We minimize the use of computationally expensive and memory-consuming layers such as triangle attention, allowing Proteina to generate backbones of up to 800 residues. (5/n)

110

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸The fold class conditioning provides fine control during generation and allows us to guide with respect to high-level secondary structure content or low-level specific fold classes. The method can also be used to enhance the amount of beta sheets in a controlled manner. (4/n)

110

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸We train on synthetic datasets of up to 21M protein structures curated from the AlphaFold Database (left plot). Further, we condition Proteina on hierarchical C.A.T.H protein structure classification labels (right plot), with a tailored classifier-free guidance scheme. (3/n)

141

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

🔸Proteina is a novel flow-based protein backbone generative model. It uses an alpha carbon backbone representation, is trained with flow matching, relies on a scalable and efficient transformer network, and offers hierarchical fold class conditioning for enhanced control. (2/n)

130

Karsten Kreis @karstenkreis.bsky.social · 04/03/2025

📢📢 "Proteina: Scaling Flow-based Protein Structure Generative Models" #ICLR2025 (Oral Presentation) 🔥 Project page: research.nvidia.com/labs/genair/... 📜 Paper: arxiv.org/abs/2503.00710 🛠️ Code and weights: github.com/NVIDIA-Digit... 🧵Details in thread... (1/n)

13910