THFlow: A Temporally Hierarchical Flow Matching Framework for 3D Peptide Design

• URL: http://arxiv.org/abs/2502.15855v3
• Date: Sat, 01 Nov 2025 12:00:57 GMT
• Title: THFlow: A Temporally Hierarchical Flow Matching Framework for 3D Peptide Design
• Authors: Dengdeng Huang, Shikui Tu,
• Abstract summary: multimodal temporal inconsistency problem is a key factor contributing to low binding generated peptides.<n>We propose a novel flow-based generative model that explicitly models the temporal hierarchy between peptide position and conformation.<n> THFlow outperforms existing methods in generating peptides with superior stability, affinity, and diversity.
• Score: 14.436723124352817
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Deep generative models provide a promising approach to de novo 3D peptide design. Most of them jointly model the distributions of peptide's position, orientation, and conformation, attempting to simultaneously converge to the target pocket. However, in the early stage of docking, optimizing conformation-only modalities such as rotation and torsion can be physically meaningless, as the peptide is initialized far from the protein pocket and no interaction field is present. We define this problem as the multimodal temporal inconsistency problem and claim it is a key factor contributing to low binding affinity in generated peptides. To address this challenge, we propose THFlow, a novel flow matching-based multimodal generative model that explicitly models the temporal hierarchy between peptide position and conformation. It employs a polynomial based conditional flow to accelerate positional convergence early on, and later aligns it with rotation and torsion for coordinated conformation refinement under the emerging interaction field. Additionally, we incorporate interaction-related features, such as polarity, to further enhance the model's understanding of peptide-protein binding. Extensive experiments demonstrate that THFlow outperforms existing methods in generating peptides with superior stability, affinity, and diversity, offering an effective and accurate solution for advancing peptide-based therapeutic development.

Related papers

• MuCO: Generative Peptide Cyclization Empowered by Multi-stage Conformation Optimization [30.75292632688159]
  We propose a generative peptide cyclization method that models the distribution of cyclic peptide conformations conditioned on the corresponding linear peptide.<n>MuCO decouples the peptide cyclization task into three stages: topology-aware backbone design, generative side-chain packing, and physics-aware all-atom optimization.<n> Experiments on the large-scale CPSea dataset demonstrate that MuCO significantly outperforms state-of-the-art methods in consistently in physical stability, structural diversity, secondary structure recovery, and computational efficiency.
  arXiv  Detail & Related papers  (2026-01-30T10:02:15Z)
• Surface-based Molecular Design with Multi-modal Flow Matching [64.00572241268597]
  SurfFlow is a novel surface-based generative algorithm that enables comprehensive co-design of sequence, structure, and surface for peptides.<n> evaluated on the comprehensive PepMerge benchmark, SurfFlow consistently outperforms full-atom baselines across all metrics.
  arXiv  Detail & Related papers  (2026-01-08T02:19:29Z)
• FoundationSLAM: Unleashing the Power of Depth Foundation Models for End-to-End Dense Visual SLAM [50.9765003472032]
  FoundationSLAM is a learning-based monocular dense SLAM system for accurate and robust tracking and mapping.<n>Our core idea is to bridge flow estimation with reasoning by leveraging the guidance from foundation depth models.
  arXiv  Detail & Related papers  (2025-12-31T17:57:45Z)
• Generative design and validation of therapeutic peptides for glioblastoma based on a potential target ATP5A [13.133911733817563]
  We present a framework combining generative modeling and experimental validation to optimize peptides targeting ATP5A.<n>As the first lead peptide-conditioned flow matching model, POTFlow holds strong potential as a generalizable framework for therapeutic peptide design.
  arXiv  Detail & Related papers  (2025-11-19T11:17:17Z)
• Full-Atom Peptide Design via Riemannian-Euclidean Bayesian Flow Networks [16.11525464623603]
  PepBFN is the first Bayesian flow network for full atom peptide design that directly models parameter distributions in fully continuous space.<n>Specifically, PepBFN models discrete residue types by learning their continuous parameter distributions, enabling joint and smooth Bayesian updates.<n>Experiments on side chain packing, reverse folding, and binder design tasks demonstrate the strong potential of PepBFN in computational peptide design.
  arXiv  Detail & Related papers  (2025-11-18T14:13:23Z)
• Zero-Shot Cyclic Peptide Design via Composable Geometric Constraints [65.77915791312634]
  We propose CP-Composer, a novel generative framework that enables zero-shot cyclic peptide generation.<n>Our approach decomposes complex cyclization patterns into unit constraints, which are incorporated into a diffusion model.<n>Our model, despite trained with linear peptides, is capable of generating diverse target-binding cyclic peptides, reaching success rates from 38% to 84%.
  arXiv  Detail & Related papers  (2025-07-06T03:30:45Z)
• Iterative Distillation for Reward-Guided Fine-Tuning of Diffusion Models in Biomolecular Design [53.93023688824764]
  We address the problem of fine-tuning diffusion models for reward-guided generation in biomolecular design.<n>We propose an iterative distillation-based fine-tuning framework that enables diffusion models to optimize for arbitrary reward functions.<n>Our off-policy formulation, combined with KL divergence minimization, enhances training stability and sample efficiency compared to existing RL-based methods.
  arXiv  Detail & Related papers  (2025-07-01T05:55:28Z)
• CreoPep: A Universal Deep Learning Framework for Target-Specific Peptide Design and Optimization [19.795752582745397]
  Target-specific peptides, such as conotoxins, exhibit exceptional binding affinity and selectivity toward ion channels and receptors.<n>Here, we present CreoPep, a deep learning-based conditional generative framework that integrates masked language modeling with a progressive masking scheme to design high-affinity peptide mutants.<n>We validate this approach by designing conotoxin inhibitors targeting the $alpha$7 nicotinic acetylcholine receptor, achieving submicromolar potency in electrophysiological tests.
  arXiv  Detail & Related papers  (2025-05-05T15:56:39Z)
• Efficient optimization and conceptual barriers in variational finite Projected Entangled-Pair States [0.0]
  Projected entangled pair states (PEPS) on finite two-dimensional lattices are a natural ansatz for representing ground states of local many-body Hamiltonians. We propose the optimization of PEPS via an improved formulation of the time-dependent variational principle (TDVP) We demonstrate our approach's capability to naturally handle long-range interactions by exploring the phase diagram of Rydberg atom arrays with long-range interactions.
  arXiv  Detail & Related papers  (2025-03-16T16:06:44Z)
• Integrating Protein Dynamics into Structure-Based Drug Design via Full-Atom Stochastic Flows [29.49146207945794]
  Traditional structure-based drug design (SBDD) approaches typically target binding sites with rigid structures. We propose to use generative modeling for SBDD considering conformational changes of protein pockets. We show that DynamicFlow learns to transform apo pockets and noisy pockets into holo pockets and corresponding 3D molecules.
  arXiv  Detail & Related papers  (2025-03-06T00:34:44Z)
• PDeepPP:A Deep learning framework with Pretrained Protein language for peptide classification [6.55419985735241]
  We propose a deep learning framework that integrates pretrained protein language models with a neural network combining transformer and CNN for peptide classification.<n>This framework was applied to multiple tasks involving PTM site and bioactive peptide prediction, utilizing large-scale datasets to enhance the model's robustness.
  arXiv  Detail & Related papers  (2025-02-21T17:31:22Z)
• PepTune: De Novo Generation of Therapeutic Peptides with Multi-Objective-Guided Discrete Diffusion [2.6668932659159905]
  We present PepTune, a multi-objective discrete diffusion model for simultaneous generation and optimization of therapeutic peptide SMILES.<n>To guide the diffusion process, we introduce Monte Carlo Tree Guidance (MCTG), an inference-time multi-objective guidance algorithm.<n>Using PepTune, we generate diverse, chemically-modified peptides simultaneously optimized for multiple therapeutic properties.
  arXiv  Detail & Related papers  (2024-12-23T18:38:49Z)
• Structure Language Models for Protein Conformation Generation [66.42864253026053]
  Traditional physics-based simulation methods often struggle with sampling equilibrium conformations. Deep generative models have shown promise in generating protein conformations as a more efficient alternative. We introduce Structure Language Modeling as a novel framework for efficient protein conformation generation.
  arXiv  Detail & Related papers  (2024-10-24T03:38:51Z)
• Full-Atom Peptide Design based on Multi-modal Flow Matching [32.58558711545861]
  We present PepFlow, the first multi-modal deep generative model grounded in the flow-matching framework for the design of full-atom peptides. We characterize the peptide structure using rigid backbone frames within the $mathrmSE(3)$ manifold and side-chain angles on high-dimensional tori. Our approach adeptly tackles various tasks such as fix-backbone sequence design and side-chain packing through partial sampling.
  arXiv  Detail & Related papers  (2024-06-02T12:59:54Z)
• PPFlow: Target-aware Peptide Design with Torsional Flow Matching [52.567714059931646]
  We propose a target-aware peptide design method called textscPPFlow to model the internal geometries of torsion angles for the peptide structure design.<n>Besides, we establish a protein-peptide binding dataset named PPBench2024 to fill the void of massive data for the task of structure-based peptide drug design.
  arXiv  Detail & Related papers  (2024-03-05T13:26:42Z)
• Re-Dock: Towards Flexible and Realistic Molecular Docking with Diffusion Bridge [69.80471117520719]
  Re-Dock is a novel diffusion bridge generative model extended to geometric manifold. We propose energy-to-geometry mapping inspired by the Newton-Euler equation to co-model the binding energy and conformations. Experiments on designed benchmark datasets including apo-dock and cross-dock demonstrate our model's superior effectiveness and efficiency over current methods.
  arXiv  Detail & Related papers  (2024-02-18T05:04:50Z)
• AlphaFold Meets Flow Matching for Generating Protein Ensembles [11.1639408863378]
  We develop a flow-based generative modeling approach for learning and sampling the conformational landscapes of proteins. Our method provides a superior combination of precision and diversity compared to AlphaFold with MSA subsampling. Our method can diversify a static PDB structure with faster wall-clock convergence to certain equilibrium properties than replicate MD trajectories.
  arXiv  Detail & Related papers  (2024-02-07T13:44:47Z)
• Equivariant Flow Matching with Hybrid Probability Transport [69.11915545210393]
  Diffusion Models (DMs) have demonstrated effectiveness in generating feature-rich geometries. DMs typically suffer from unstable probability dynamics with inefficient sampling speed. We introduce geometric flow matching, which enjoys the advantages of both equivariant modeling and stabilized probability dynamics.
  arXiv  Detail & Related papers  (2023-12-12T11:13:13Z)
• Harmonic Self-Conditioned Flow Matching for Multi-Ligand Docking and Binding Site Design [29.972698307121675]
  FlowSite extends this flow model to jointly generate a protein pocket's discrete residue types and the molecule's binding 3D structure. We show HarmonicFlow improves upon state-of-the-art generative processes for docking in simplicity, generality, and average sample quality.
  arXiv  Detail & Related papers  (2023-10-09T14:45:33Z)
• SE(3)-Stochastic Flow Matching for Protein Backbone Generation [54.951832422425454]
  We introduce FoldFlow, a series of novel generative models of increasing modeling power based on the flow-matching paradigm over $3mathrmD$ rigid motions. Our family of FoldFlowgenerative models offers several advantages over previous approaches to the generative modeling of proteins.
  arXiv  Detail & Related papers  (2023-10-03T19:24:24Z)
• Protein Design with Guided Discrete Diffusion [67.06148688398677]
  A popular approach to protein design is to combine a generative model with a discriminative model for conditional sampling. We propose diffusioN Optimized Sampling (NOS), a guidance method for discrete diffusion models. NOS makes it possible to perform design directly in sequence space, circumventing significant limitations of structure-based methods.
  arXiv  Detail & Related papers  (2023-05-31T16:31:24Z)
• EBM-Fold: Fully-Differentiable Protein Folding Powered by Energy-based Models [53.17320541056843]
  We propose a fully-differentiable approach for protein structure optimization, guided by a data-driven generative network. Our EBM-Fold approach can efficiently produce high-quality decoys, compared against traditional Rosetta-based structure optimization routines.
  arXiv  Detail & Related papers  (2021-05-11T03:40:29Z)
• Designing a Prospective COVID-19 Therapeutic with Reinforcement Learning [50.57291257437373]
  SARS-CoV-2 pandemic has created a global race for a cure. One approach focuses on designing a novel variant of the human angiotensin-converting enzyme 2 (ACE2) We formulate a novel protein design framework as a reinforcement learning problem.
  arXiv  Detail & Related papers  (2020-12-03T07:35:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.