pith. sign in

arxiv: 2605.23407 · v1 · pith:3QMVPV4Onew · submitted 2026-05-22 · 💻 cs.CE

GeoCycler: Reward-Aligned 3D Diffusion for Constraint-Conditioned Cyclic Peptide Design

Jingjie Zhang , Hanqun Cao , Haosen Shi , He Mutian , Yu Wang , Zijun Gao , Fang Wu , Xiaojun Yao
show 5 more authors
Chang-Yu Hsieh Sinno Jialin Pan Pranam Chatterjee Chunbin Gu Pheng-Ann Heng
This is my paper

Pith reviewed 2026-05-25 02:43 UTC · model grok-4.3

classification 💻 cs.CE
keywords cyclic peptidesdiffusion modelsreward alignmentmacrocyclizationgeometric constraintspeptide designconstraint-conditioned generation3D structure generation
0
0 comments X

The pith

Training a diffusion model with selective rewards at generation time improves cyclic peptide closure success over post-generation guidance.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Cyclic peptides need closed-ring structures for stability and specificity, but diffusion generators often fail to meet the required geometric constraints during sampling. The paper argues that reshaping the model's learned distribution through training-time reward alignment produces more valid closed structures than steering samples only at inference time. It introduces a reward signal that activates geometric penalties selectively based on residue types, paired with positive weighting and replay buffers to handle multiple closure topologies in one model. Experiments on the LNR benchmark show higher rates of successful closures across stapled, head-to-tail, disulfide, and bicyclic cases, with a large gain for head-to-tail cyclization and no major shift in amino-acid or dihedral distributions. This positions training-time alignment as a direct way to embed sparse constraints into the generator rather than correcting outputs afterward.

Core claim

GeoCycler aligns a single generator across multiple cyclization topologies by introducing a type-gated stair reward that activates distance-based shaping only when prerequisite residue or linker types are satisfied, together with positive-only reward weighting and replay-based stabilization, resulting in improved pass@5 closure success on the LNR benchmark, including a 20.8 percentage point gain in head-to-tail success over CP-Composer while maintaining comparable amino-acid and backbone-dihedral statistics.

What carries the argument

The type-gated stair reward inside a reward-weighted diffusion alignment framework for conditional latent diffusion models, which supplies dense geometric feedback only for chemically compatible anchors to reshape the generative distribution toward macrocyclization feasibility.

If this is right

  • A single trained model achieves higher closure success across stapled, head-to-tail, disulfide, and bicyclic settings without separate guidance schedules.
  • Head-to-tail closure success rises by 20.8 percentage points over CP-Composer on the LNR benchmark.
  • Amino-acid composition and backbone dihedral statistics remain comparable to unaligned baselines.
  • Training-time alignment serves as an alternative to relying solely on inference-time correction for sparse geometric constraints.
  • The framework supports alignment across multiple cyclization topologies in one generator.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The selective reward approach could transfer to other 3D generative tasks with sparse contact constraints, such as designing proteins with specific disulfide patterns.
  • Combining the alignment with additional property rewards might enable multi-objective peptide design without separate sampling stages.
  • If the type-gating logic generalizes, similar methods could stabilize training for macrocyclic small molecules beyond peptides.
  • Efficiency gains in design pipelines could arise from fewer rejected samples, though this depends on whether diversity holds at scale.

Load-bearing premise

The type-gated stair reward combined with positive-only weighting and replay stabilization can reshape the learned generative distribution to satisfy sparse macrocyclization constraints without introducing new biases or reducing sample diversity across the four topologies.

What would settle it

An evaluation on the LNR benchmark showing that GeoCycler produces no higher pass@5 closure success than strong guidance baselines on head-to-tail or other topologies, or that amino-acid and dihedral statistics diverge markedly, would falsify the claim.

Figures

Figures reproduced from arXiv: 2605.23407 by Chang-Yu Hsieh, Chunbin Gu, Fang Wu, Hanqun Cao, Haosen Shi, He Mutian, Jingjie Zhang, Pheng-Ann Heng, Pranam Chatterjee, Sinno Jialin Pan, Xiaojun Yao, Yu Wang, Zijun Gao.

Figure 1
Figure 1. Figure 1: Cyclic peptide generation under hybrid geometric constraints. Cyclic peptide design requires satisfying both discrete prerequisite conditions, such as compatible anchor residues or linker motifs, and continuous closure-oriented geometry. GeoCycler studies whether training-time policy alignment can increase the probability of generating closure-consistent candidates, while post-hoc structural screens are us… view at source ↗
Figure 2
Figure 2. Figure 2: Overview of GeoCycler. GeoCycler fine-tunes a conditional latent diffusion generator through reward-weighted policy alignment. The framework combines type-gated geometric surrogate rewards, positive-only updates, and replay-based stabilization to increase the probability of closure￾consistent cyclic peptide candidates across multiple topology constraints. reward-aligned baselines, with particularly strong … view at source ↗
Figure 3
Figure 3. Figure 3: Representative structural realizations across four macrocyclization topologies. Panels a–d show GeoCycler samples, and panels e–h show CP-Composer samples. Red boxes highlight the local regions associated with cyclization constraints. 4.4 Qualitative Structural Realization [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
read the original abstract

Cyclic peptides are attractive therapeutic modalities because their closed-ring topology can improve stability and target specificity. However, de novo cyclic peptide design remains challenging for diffusion generators, as macrocyclization requires satisfying sparse, non-smooth, and compositional geometric constraints. Existing constraint-conditioned methods largely rely on inference-time guidance, which can steer samples toward desired closures but does not directly change the learned generative distribution. We propose GeoCycler, a reward-weighted diffusion alignment framework for training conditional latent diffusion models toward macrocyclization feasibility. GeoCycler introduces a type-gated stair reward that activates distance-based shaping only when prerequisite residue or linker types are satisfied, providing dense geometric feedback while avoiding misleading signals from chemically incompatible anchors. Together with positive-only reward weighting and replay-based stabilization, GeoCycler aligns a single generator across multiple cyclization topologies. On the LNR benchmark, GeoCycler improves pass@5 closure success over strong guidance-based baselines across stapled, head-to-tail, disulfide, and bicyclic settings. In particular, it improves head-to-tail success by 20.8 percentage points over CP-Composer while maintaining comparable amino-acid and backbone-dihedral statistics. These results suggest that training-time alignment to sparse geometric constraints is a promising alternative to relying solely on post hoc sampling-time correction for cyclic peptide generation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 0 minor

Summary. The paper proposes GeoCycler, a reward-weighted diffusion alignment framework for training conditional latent diffusion models to generate cyclic peptides satisfying macrocyclization constraints. It introduces a type-gated stair reward that provides dense geometric feedback only when residue or linker types are compatible, combined with positive-only reward weighting and replay-based stabilization to align a single generator across stapled, head-to-tail, disulfide, and bicyclic topologies. On the LNR benchmark, the method is claimed to improve pass@5 closure success over guidance-based baselines, including a 20.8 percentage point gain on head-to-tail closure relative to CP-Composer, while preserving comparable amino-acid composition and backbone-dihedral statistics.

Significance. If the empirical results hold after proper controls, the work would indicate that training-time reward alignment can reshape the generative distribution of 3D diffusion models to satisfy sparse, non-smooth geometric constraints more effectively than inference-time guidance alone. This would be relevant to computational peptide design, as it offers a mechanism for handling compositional cyclization requirements without post-hoc correction.

major comments (1)
  1. [Abstract] Abstract: the central empirical claim of a 20.8 pp improvement in head-to-tail pass@5 success (and gains across four topologies) is presented without any description of experimental controls, number of samples, error bars, data splits, statistical tests, or baseline implementation details, rendering the quantitative result unverifiable from the provided text.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their review and for identifying the lack of experimental context in the abstract. We address this point directly below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central empirical claim of a 20.8 pp improvement in head-to-tail pass@5 success (and gains across four topologies) is presented without any description of experimental controls, number of samples, error bars, data splits, statistical tests, or baseline implementation details, rendering the quantitative result unverifiable from the provided text.

    Authors: We agree the abstract omits these details. The main manuscript (Section 4) specifies 1000 samples per method per topology, 5 independent seeds for reporting means and standard deviations, the standard LNR train/test splits, and baseline re-implementations matching the original CP-Composer settings; statistical comparisons appear in the supplement. We will revise the abstract to include a concise clause such as 'across 1000 samples per topology with 5 seeds' while preserving length, and will add a pointer to the methods for full controls. This change will appear in the next version. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper describes an empirical ML framework (reward-weighted diffusion alignment with type-gated stair rewards, positive-only weighting, and replay stabilization) evaluated on the LNR benchmark for cyclic peptide closure success rates. All load-bearing claims are experimental outcomes (e.g., +20.8 pp head-to-tail pass@5 improvement) rather than mathematical derivations, first-principles predictions, or quantities defined in terms of themselves. No equations reduce to self-definitions, no fitted parameters are relabeled as predictions, and no self-citation chain is invoked to justify uniqueness or force the central result. The method is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no specific free parameters, axioms, or invented entities can be identified from the provided text.

pith-pipeline@v0.9.0 · 5816 in / 1220 out tokens · 27815 ms · 2026-05-25T02:43:43.484350+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

70 extracted references · 70 canonical work pages · 4 internal anchors

  1. [1]

    Current opinion in chemical biology , volume=

    Cyclic peptide therapeutics: past, present and future , author=. Current opinion in chemical biology , volume=. 2017 , publisher=

    work page 2017

  2. [2]

    Forty-second International Conference on Machine Learning , year=

    PepTune: De Novo Generation of Therapeutic Peptides with Multi-Objective-Guided Discrete Diffusion , author=. Forty-second International Conference on Machine Learning , year=

    work page

  3. [5]

    ICML 2025 Generative AI and Biology (GenBio) Workshop , year=

    SOAPIA: Siamese-guided generation of off target-avoiding protein interactions with high target affinity , author=. ICML 2025 Generative AI and Biology (GenBio) Workshop , year=

    work page 2025

  4. [6]

    ICML 2025 Generative AI and Biology (GenBio) Workshop , year=

    Multi-Objective-Guided Discrete Flow Matching for Controllable Biological Sequence Design , author=. ICML 2025 Generative AI and Biology (GenBio) Workshop , year=

    work page 2025

  5. [7]

    Nature Methods , pages=

    PTM-Mamba: a PTM-aware protein language model with bidirectional gated Mamba blocks , author=. Nature Methods , pages=. 2025 , publisher=

    work page 2025

  6. [8]

    Science Advances , volume=

    De novo design of peptide binders to conformationally diverse targets with contrastive language modeling , author=. Science Advances , volume=. 2025 , publisher=

    work page 2025

  7. [9]

    Nature Biotechnology , pages=

    Target sequence-conditioned design of peptide binders using masked language modeling , author=. Nature Biotechnology , pages=. 2025 , publisher=

    work page 2025

  8. [10]

    Signal transduction and targeted therapy , volume=

    Therapeutic peptides: current applications and future directions , author=. Signal transduction and targeted therapy , volume=. 2022 , publisher=

    work page 2022

  9. [11]

    moPPIt-v3: Motif-specific peptides generated via multi-objective-guided discrete flow matching , author=

    work page

  10. [12]

    Gumbel-Softmax Score and Flow Matching for Discrete Biological Sequence Generation , author=

    work page

  11. [13]

    Science , volume=

    Robust deep learning--based protein sequence design using ProteinMPNN , author=. Science , volume=. 2022 , publisher=

    work page 2022

  12. [14]

    Nature Chemical Biology , pages=

    Accurate de novo design of high-affinity protein-binding macrocycles using deep learning , author=. Nature Chemical Biology , pages=. 2025 , publisher=

    work page 2025

  13. [16]

    Designing Cyclic Peptides via Harmonic

    Xiangxin Zhou and Mingyu Li and Yi Xiao and Jiahan Li and Dongyu Xue and Zaixiang Zheng and Jianzhu Ma and Quanquan Gu , booktitle=. Designing Cyclic Peptides via Harmonic. 2025 , url=

    work page 2025

  14. [17]

    Advances in Neural Information Processing Systems , volume=

    Full-atom peptide design with geometric latent diffusion , author=. Advances in Neural Information Processing Systems , volume=

    work page

  15. [18]

    The Eleventh International Conference on Learning Representations , year=

    Diffusion Posterior Sampling for General Noisy Inverse Problems , author=. The Eleventh International Conference on Learning Representations , year=

    work page

  16. [20]

    arXiv preprint arXiv:2206.04119 , year=

    Diffusion probabilistic modeling of protein backbones in 3d for the motif-scaffolding problem , author=. arXiv preprint arXiv:2206.04119 , year=

    work page arXiv

  17. [21]

    nature , volume=

    Highly accurate protein structure prediction with AlphaFold , author=. nature , volume=. 2021 , publisher=

    work page 2021

  18. [22]

    Journal of Medicinal Chemistry , year=

    HighPlay: Cyclic Peptide Sequence Design Based on Reinforcement Learning and Protein Structure Prediction , author=. Journal of Medicinal Chemistry , year=

    work page

  19. [23]

    Nature , volume=

    De novo design of protein structure and function with RFdiffusion , author=. Nature , volume=. 2023 , publisher=

    work page 2023

  20. [24]

    International conference on machine learning , pages=

    Equivariant diffusion for molecule generation in 3d , author=. International conference on machine learning , pages=. 2022 , organization=

    work page 2022

  21. [25]

    Nature Communications , volume=

    Cyclic peptide structure prediction and design using AlphaFold2 , author=. Nature Communications , volume=. 2025 , publisher=

    work page 2025

  22. [27]

    bioRxiv , year=

    A deep reinforcement learning platform for antibiotic discovery , author=. bioRxiv , year=

    work page

  23. [28]

    Nature Machine Intelligence , volume=

    Self-play reinforcement learning guides protein engineering , author=. Nature Machine Intelligence , volume=. 2023 , publisher=

    work page 2023

  24. [30]

    2025 , url=

    Hanqun Cao and Haosen Shi and Chenyu Wang and Sinno Jialin Pan and Pheng-Ann Heng , booktitle=. 2025 , url=

    work page 2025

  25. [31]

    Fine-Tuning Discrete Diffusion Models via Reward Optimization with Applications to DNA and Protein Design , author=

    work page

  26. [32]

    Science , volume=

    Top-down design of protein architectures with reinforcement learning , author=. Science , volume=. 2023 , publisher=

    work page 2023

  27. [35]

    Journal of Medicinal Chemistry , year=

    Reinforcement Learning-Based Target-Specific De Novo Design of Cyclic Peptide Binders , author=. Journal of Medicinal Chemistry , year=

    work page

  28. [37]

    Nature Machine Intelligence , pages=

    Accelerating protein engineering with fitness landscape modelling and reinforcement learning , author=. Nature Machine Intelligence , pages=. 2025 , publisher=

    work page 2025

  29. [38]

    Angewandte Chemie , volume=

    Cyclic peptides for drug development , author=. Angewandte Chemie , volume=. 2024 , publisher=

    work page 2024

  30. [39]

    Chemical Reviews , volume=

    Understanding cell penetration of cyclic peptides , author=. Chemical Reviews , volume=. 2019 , publisher=

    work page 2019

  31. [40]

    Nucleic acids research , volume=

    Rosetta FlexPepDock web server—high resolution modeling of peptide--protein interactions , author=. Nucleic acids research , volume=. 2011 , publisher=

    work page 2011

  32. [43]

    Auto-Encoding Variational Bayes

    Auto-encoding variational bayes , author=. arXiv preprint arXiv:1312.6114 , year=

    work page internal anchor Pith review Pith/arXiv arXiv

  33. [45]

    Advances in Neural Information Processing Systems , volume=

    Dpok: Reinforcement learning for fine-tuning text-to-image diffusion models , author=. Advances in Neural Information Processing Systems , volume=

    work page

  34. [46]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Diffusion model alignment using direct preference optimization , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    work page

  35. [47]

    Nature communications , volume=

    Harnessing protein folding neural networks for peptide--protein docking , author=. Nature communications , volume=. 2022 , publisher=

    work page 2022

  36. [49]

    F. Bao, M. Zhao, Z. Hao, P. Li, C. Li, and J. Zhu. Equivariant energy-guided sde for inverse molecular design. arXiv preprint arXiv:2209.15408, 2022

    work page arXiv 2022

  37. [50]

    Training Diffusion Models with Reinforcement Learning

    K. Black, M. Janner, Y. Du, I. Kostrikov, and S. Levine. Training diffusion models with reinforcement learning. arXiv preprint arXiv:2305.13301, 2023

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  38. [51]

    H. Cao, H. Shi, C. Wang, S. J. Pan, and P.-A. Heng. GLID \ 2\ e: A gradient-free lightweight fine-tune approach for discrete biological sequence design. In The Thirty-ninth Annual Conference on Neural Information Processing Systems, 2025 a . URL https://openreview.net/forum?id=AHjspi4R22

    work page 2025

  39. [52]

    H. Cao, M. D. Torres, J. Zhang, Z. Gao, F. Wu, C. Gu, J. Leskovec, Y. Choi, C. de la Fuente-Nunez, G. Chen, et al. A deep reinforcement learning platform for antibiotic discovery. bioRxiv, 2025 b

    work page 2025

  40. [53]

    H. Cao, H. Zhang, J. Xu, Z. Zhang, L. Shen, M. Sun, G. Liu, J. Xu, W.-J. Li, J. Ni, et al. From supervision to exploration: What does protein language model learn during reinforcement learning? arXiv preprint arXiv:2510.01571, 2025 c

    work page arXiv 2025

  41. [54]

    T. Chen, Y. Zhang, and P. Chatterjee. Areuredi: Annealed rectified updates for refining discrete flows with multi-objective guidance. arXiv preprint arXiv:2510.00352, 2025 a

    work page arXiv 2025

  42. [55]

    T. Chen, Y. Zhang, S. Tang, and P. Chatterjee. Multi-objective-guided discrete flow matching for controllable biological sequence design. In ICML 2025 Generative AI and Biology (GenBio) Workshop, 2025 b . URL https://openreview.net/forum?id=8YIMLoHP9J

    work page 2025

  43. [56]

    P. G. Dougherty, A. Sahni, and D. Pei. Understanding cell penetration of cyclic peptides. Chemical Reviews, 119 0 (17): 0 10241--10287, 2019

    work page 2019

  44. [57]

    Y. Fan, O. Watkins, Y. Du, H. Liu, M. Ryu, C. Boutilier, P. Abbeel, M. Ghavamzadeh, K. Lee, and K. Lee. Dpok: Reinforcement learning for fine-tuning text-to-image diffusion models. Advances in Neural Information Processing Systems, 36: 0 79858--79885, 2023

    work page 2023

  45. [58]

    Z. Gao, T. Feng, J. You, C. Zi, Y. Zhou, C. Zhang, and J. Li. Deep reinforcement learning for modelling protein complexes. arXiv preprint arXiv:2405.02299, 2024

    work page arXiv 2024

  46. [59]

    Classifier-Free Diffusion Guidance

    J. Ho and T. Salimans. Classifier-free diffusion guidance. arXiv preprint arXiv:2207.12598, 2022

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  47. [60]

    Hoogeboom, V

    E. Hoogeboom, V. G. Satorras, C. Vignac, and M. Welling. Equivariant diffusion for molecule generation in 3d. In International conference on machine learning, pages 8867--8887. PMLR, 2022

    work page 2022

  48. [61]

    X. Ji, A. L. Nielsen, and C. Heinis. Cyclic peptides for drug development. Angewandte Chemie, 136 0 (3): 0 e202308251, 2024

    work page 2024

  49. [62]

    Jiang, X

    D. Jiang, X. Kong, J. Han, M. Li, R. Jiao, W. Huang, S. Ermon, J. Ma, and Y. Liu. Zero-shot cyclic peptide design via composable geometric constraints. arXiv preprint arXiv:2507.04225, 2025

    work page arXiv 2025

  50. [63]

    Jumper, R

    J. Jumper, R. Evans, A. Pritzel, T. Green, M. Figurnov, O. Ronneberger, K. Tunyasuvunakool, R. Bates, A. Z \' dek, A. Potapenko, et al. Highly accurate protein structure prediction with alphafold. nature, 596 0 (7873): 0 583--589, 2021

    work page 2021

  51. [64]

    X. Kong, Y. Jia, W. Huang, and Y. Liu. Full-atom peptide design with geometric latent diffusion. Advances in Neural Information Processing Systems, 37: 0 74808--74839, 2024

    work page 2024

  52. [65]

    J. Li, C. Cheng, Z. Wu, R. Guo, S. Luo, Z. Ren, J. Peng, and J. Ma. Full-atom peptide design based on multi-modal flow matching. arXiv preprint arXiv:2406.00735, 2024

    work page arXiv 2024

  53. [66]

    H. Lin, C. Zhu, T. Shang, N. Zhu, K. Lin, C. Zhang, X. Shao, X. Wang, and H. Duan. Highplay: Cyclic peptide sequence design based on reinforcement learning and protein structure prediction. Journal of Medicinal Chemistry, 2025

    work page 2025

  54. [67]

    M. Liu, X. Cheng, Z. Gao, H. Chang, C. Tan, S. Shan, and X. Chen. Protinvtree: Deliberate protein inverse folding with reward-guided tree search. arXiv preprint arXiv:2506.00925, 2025

    work page arXiv 2025

  55. [68]

    I. D. Lutz, S. Wang, C. Norn, A. Courbet, A. J. Borst, Y. T. Zhao, A. Dosey, L. Cao, J. Xu, E. M. Leaf, et al. Top-down design of protein architectures with reinforcement learning. Science, 380 0 (6642): 0 266--273, 2023

    work page 2023

  56. [69]

    S. A. Rettie, K. V. Campbell, A. K. Bera, A. Kang, S. Kozlov, Y. F. Bueso, J. De La Cruz, M. Ahlrichs, S. Cheng, S. R. Gerben, et al. Cyclic peptide structure prediction and design using alphafold2. Nature Communications, 16 0 (1): 0 4730, 2025 a

    work page 2025

  57. [70]

    S. A. Rettie, D. Juergens, V. Adebomi, Y. F. Bueso, Q. Zhao, A. N. Leveille, A. Liu, A. K. Bera, J. A. Wilms, A. \"U ffing, et al. Accurate de novo design of high-affinity protein-binding macrocycles using deep learning. Nature Chemical Biology, pages 1--9, 2025 b

    work page 2025

  58. [71]

    H. Sun, L. He, P. Deng, G. Liu, Z. Zhao, Y. Jiang, C. Cao, F. Ju, L. Wu, H. Liu, et al. Accelerating protein engineering with fitness landscape modelling and reinforcement learning. Nature Machine Intelligence, pages 1--15, 2025

    work page 2025

  59. [72]

    S. Tang, Y. Zhang, and P. Chatterjee. Peptune: De novo generation of therapeutic peptides with multi-objective-guided discrete diffusion. In Forty-second International Conference on Machine Learning, 2025 a . URL https://openreview.net/forum?id=FQoy1Y1Hd8

    work page 2025

  60. [73]

    S. Tang, Y. Zhu, M. Tao, and P. Chatterjee. Tr2-d2: Tree search guided trajectory-aware fine-tuning for discrete diffusion. arXiv preprint arXiv:2509.25171, 2025 b

    work page arXiv 2025

  61. [74]

    Tsaban, J

    T. Tsaban, J. K. Varga, O. Avraham, Z. Ben-Aharon, A. Khramushin, and O. Schueler-Furman. Harnessing protein folding neural networks for peptide--protein docking. Nature communications, 13 0 (1): 0 176, 2022

    work page 2022

  62. [75]

    Wallace, M

    B. Wallace, M. Dang, R. Rafailov, L. Zhou, A. Lou, S. Purushwalkam, S. Ermon, C. Xiong, S. Joty, and N. Naik. Diffusion model alignment using direct preference optimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8228--8238, 2024

    work page 2024

  63. [76]

    F. Wang, T. Zhang, J. Zhu, X. Zhang, C. Zhang, and L. Lai. Reinforcement learning-based target-specific de novo design of cyclic peptide binders. Journal of Medicinal Chemistry, 2025

    work page 2025

  64. [77]

    L. Wang, N. Wang, W. Zhang, X. Cheng, Z. Yan, G. Shao, X. Wang, R. Wang, and C. Fu. Therapeutic peptides: current applications and future directions. Signal transduction and targeted therapy, 7 0 (1): 0 48, 2022

    work page 2022

  65. [78]

    Y. Wang, H. Tang, L. Huang, L. Pan, L. Yang, H. Yang, F. Mu, and M. Yang. Self-play reinforcement learning guides protein engineering. Nature Machine Intelligence, 5 0 (8): 0 845--860, 2023

    work page 2023

  66. [79]

    J. L. Watson, D. Juergens, N. R. Bennett, B. L. Trippe, J. Yim, H. E. Eisenach, W. Ahern, A. J. Borst, R. J. Ragotte, L. F. Milles, et al. De novo design of protein structure and function with rfdiffusion. Nature, 620 0 (7976): 0 1089--1100, 2023

    work page 2023

  67. [80]

    J. Xu, Z. Gao, X. Zhou, J. Hu, X. Cheng, L. Song, G. Chen, P.-A. Heng, and J. Qiu. Protein inverse folding from structure feedback. arXiv preprint arXiv:2506.03028, 2025

    work page arXiv 2025

  68. [81]

    DiffusionNFT: Online Diffusion Reinforcement with Forward Process

    K. Zheng, H. Chen, H. Ye, H. Wang, Q. Zhang, K. Jiang, H. Su, S. Ermon, J. Zhu, and M.-Y. Liu. Diffusionnft: Online diffusion reinforcement with forward process. arXiv preprint arXiv:2509.16117, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  69. [82]

    X. Zhou, M. Li, Y. Xiao, J. Li, D. Xue, Z. Zheng, J. Ma, and Q. Gu. Designing cyclic peptides via harmonic SDE with atom-bond modeling. In Forty-second International Conference on Machine Learning, 2025. URL https://openreview.net/forum?id=ERu2ZiAnR7

    work page 2025

  70. [83]

    Zorzi, K

    A. Zorzi, K. Deyle, and C. Heinis. Cyclic peptide therapeutics: past, present and future. Current opinion in chemical biology, 38: 0 24--29, 2017

    work page 2017