Sequence-Augmented SE(3)-Flow Matching For Conditional Protein Generation
Guillaume Huguet, James Vuckovic, Kilian Fatras, Eric Thibodeau-Laufer, Pablo Lemos, Riashat Islam, Cheng-Hao Liu, Jarrid Rector-Brooks, Tara Akhound-Sadegh, Michael M. Bronstein, Alexander Tong, Avishek Joey Bose
Abstract
Proteins are essential for almost all biological processes and derive their diverse functions from complex 3D structures, which are in turn determined by their amino acid sequences. In this paper, we exploit the rich biological inductive bias of amino acid sequences and introduce FOLDFLOW-2 4 , a novel sequence-conditioned SE(3)-equivariant flow matching model for protein structure generation. FOLDFLOW-2 presents substantial new architectural features over the previous FOLDFLOW family of models including a protein large language model to encode sequence, a new multi-modal fusion trunk that combines structure and sequence representations, and a geometric transformer based decoder. To increase diversity and novelty of generated samples-crucial for de-novo drug design-we train FOLDFLOW-2 at scale on a new dataset that is an order of magnitude larger than PDB datasets of prior works, containing both known proteins in PDB and high-quality synthetic structures achieved through filtering. We further demonstrate the ability to align FOLDFLOW-2 to arbitrary rewards, e.g. increasing secondary structures diversity, by introducing a Reinforced Finetuning (ReFT) objective. We empirically observe that FOLDFLOW-2 outperforms previous state-of-the-art protein structure-based generative models, improving over RFDiffusion in terms of unconditional generation across all metrics including designability, diversity, and novelty across all protein lengths, as well as exhibiting generalization on the task of equilibrium conformation sampling. Finally, we demonstrate that a fine-tuned FOLDFLOW-2 makes progress on challenging conditional design tasks such as designing scaffolds for the VHH nanobody.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
13 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-Design cites SE(3) diffusion model with application to protein backbone generation
- Learning conformational ensembles of proteins based on backbone geometry cites SE(3) diffusion model with application to protein backbone generation
- Finetuned Language Models are Zero-Shot Learners cites Training language models to follow instructions with human feedback
- Training language models to follow instructions with human feedback cites Finetuned Language Models are Zero-Shot Learners
- Flow Autoencoders are Effective Protein Tokenizers cites SE(3) diffusion model with application to protein backbone generation
- Training language models to follow instructions with human feedback cites Cross-Task Generalization via Natural Language Crowdsourcing Instructions
- Cross-Task Generalization via Natural Language Crowdsourcing Instructions cites Finetuned Language Models are Zero-Shot Learners
Cited by top-tier papers13
- Generalised Flow Maps for Few-Step Generative Modelling on Riemannian ManifoldsOscar Davis, Michael S. Albergo, Nicholas M. Boffi, Michael M. Bronstein et al.ICLR 2026 · 12 citations
- Riemannian Flow Matching for Brain Connectivity Matrices via Pullback GeometryAntoine Collas, Ce Ju, Nicolas Salvy, Bertrand ThirionNeurIPS 2025 · 11 citations
- Learning conformational ensembles of proteins based on backbone geometryNicolas Wolf, Leif Seute, Vsevolod Viliuga, Simon Wagner et al.NeurIPS 2025 · 7 citations
- Flow Autoencoders are Effective Protein TokenizersRohit Dilip, Evan Zhang, Ayush Varshney, David Van ValenICLR 2026 · 5 citations
- RFMPose: Generative Category-level Object Pose Estimation via Riemannian Flow MatchingWenzhe Ouyang, Qi Ye, Jinghua Wang, Zenglin Xu et al.NeurIPS 2025 · 5 citations
Builds on17
- Training language models to follow instructions with human feedbackLong Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida et al.NeurIPS 2022 · 24,707 citations
- Finetuned Language Models are Zero-Shot LearnersJason Wei, Maarten Bosma, Vincent Y. Zhao, Kelvin Guu et al.ICLR 2022 · 4,966 citations
- Cross-Task Generalization via Natural Language Crowdsourcing InstructionsSwaroop Mishra, Daniel Khashabi, Chitta Baral, Hannaneh HajishirziACL 2022 · 887 citations
- SE(3) diffusion model with application to protein backbone generationJason Yim, Brian L. Trippe, Valentin De Bortoli, Emile Mathieu et al.ICML 2023 · 313 citations
- Generative Flows on Discrete State-Spaces: Enabling Multimodal Flows with Applications to Protein Co-DesignAndrew Campbell, Jason Yim, Regina Barzilay, Tom Rainforth et al.ICML 2024 · 283 citations
Related papers
- FlexiFlow: decomposable flow matching for generation of flexible molecular ensembleRiccardo Tedoldi, Ola Engkvist, Patrick Bryant, Hossein Azizpour et al.ICML 2026
- Proteina: Scaling Flow-based Protein Structure Generative ModelsTomas Geffner, Kieran Didi, Zuobai Zhang, Danny Reidenbach et al.ICLR 2025
- Fold2Seq: A Joint Sequence(1D)-Fold(3D) Embedding-based Generative Model for Protein DesignYue Cao, Payel Das, Vijil Chenthamarakshan, Pin-Yu Chen et al.ICML 2021 · 56 citations
- La-Proteina: Atomistic Protein Generation via Partially Latent Flow MatchingTomas Geffner, Kieran Didi, Zhonglin Cao, Danny Reidenbach et al.ICLR 2026 · 57 citations
- SE(3)-Stochastic Flow Matching for Protein Backbone GenerationAvishek Joey Bose, Tara Akhound-Sadegh, Guillaume Huguet, Kilian Fatras et al.ICLR 2024 · 162 citations
