Skip to main content
QUICK REVIEW

[论文解读] Minimal-Action Discrete Schrödinger Bridge Matching for Peptide Sequence Design

Shrey Goel, Pranam Chatterjee|arXiv (Cornell University)|Jan 29, 2026
DNA and Biological Computing被引用 0
一句话总结

MadSBM 引入基于速率的离散薛定谔桥接用于肽设计,使用生物学知情的参考过程和学习得到的控制以高效地产生高似然性序列。

ABSTRACT

Generative modeling of peptide sequences requires navigating a discrete and highly constrained space in which many intermediate states are chemically implausible or unstable. Existing discrete diffusion and flow-based methods rely on reversing fixed corruption processes or following prescribed probability paths, which can force generation through low-likelihood regions and require countless sampling steps. We introduce Minimal-action discrete Schrödinger Bridge Matching (MadSBM), a rate-based generative framework for peptide design that formulates generation as a controlled continuous-time Markov process on the amino-acid edit graph. To yield probability trajectories that remain near high-likelihood sequence neighborhoods throughout generation, MadSBM 1) defines generation relative to a biologically informed reference process derived from pre-trained protein language model logits and 2) learns a time-dependent control field that biases transition rates to produce low-action transport paths from a masked prior to the data distribution. We finally introduce guidance to the MadSBM sampling procedure towards a specific functional objective, expanding the design space of therapeutic peptides; to our knowledge, this represents the first-ever application of discrete classifier guidance to Schrödinger bridge-based generative models.

研究动机与目标

  • 将肽序列生成重新表述为在氨基酸编辑图上的噪声先验与数据分布之间的最小作用传输
  • 利用来自预训练蛋白质语言模型 logits 的生物学知情参考过程来引导生成
  • 学习时变控制场以使跃迁速率偏向高似然性肽邻域
  • 实现目标导向采样以结合功能设计目标
  • 在与离散扩散基线相比时 Demonstrate 样本效率和稳定性,同时实现引导设计

提出的方法

  • 将生成建模为在氨基酸编辑图上的受控连续时间马尔可夫链(CTMC),参考生成器 R0
  • 定义 R_u(x,x') = R0(x,x') exp(u_theta(x,x',t)) 以创建倾向化、时变的受控过程
  • 推导可行的作用泛函 A(u) = E_{P_u}[ ∫ R0(X_t,x') Ψ(u(X_t,x',t)) dt ],其中 Ψ(z)=e^z - z - 1
  • 利用交叉熵目标训练神经控制场 u_theta,使跃迁朝向目标序列对齐(式 (Eq. 10))
  • 使用 ESM-2 logits 作为参考过程,并通过时间门控机制在掩蔽过程上调节参考的影响
  • 通过学习的生成器 R_theta 离散化成 N 步,对完全掩蔽的先验进行 CTMC 模拟来进行采样
Figure 1: Overview of MadSBM . We leverage a principled reference process $R_{0}$ so the MadSBM model requires only a lightweight time-conditioned control field $u_{\theta}$ to steer samples toward high-likelihood regions of the sequence space.
Figure 1: Overview of MadSBM . We leverage a principled reference process $R_{0}$ so the MadSBM model requires only a lightweight time-conditioned control field $u_{\theta}$ to steer samples toward high-likelihood regions of the sequence space.

实验结果

研究问题

  • RQ1能否在生物信息知情的参考过程下,在全新的离散肽空间中求解离散薛定谔桥?
  • RQ2在最小作用目标下学习时变控制场,是否能相较于离散扩散基线实现高效且高质量的肽生成?
  • RQ3是否可以将目标引导(如结合亲和力)纳入离散薛定谔桥采样,以提升功能设计?

主要发现

  • MadSBM 在不同采样预算下的无条件序列质量与离散扩散基线相比具有竞争力甚至更好(PPL 较低),且 pLDDT 得分合理
  • 无条件结果显示 MadSBM 的 PPL 分别为:8.389、8.943、8.719,而 DD 的 PPL 分别为:10.990、9.042、7.617,对应 N=32、64、128(越低越好)
  • MadSBM 在采样过程中呈现更广的似然路径多样性,较 DD 基线更早收敛到高似然区域
  • 消融对生物信息参考(ESM-2)和时间门控的影响会降低困惑度,验证了参考动态设计的有效性
  • 在结合结合力预测器的目标引导采样下,相较于无条件生成,能够在若干目标上提升亲和力和对接分数
Figure 2: Probability paths taken by models under various sampling budgets $(N)$ . The y-axis represents the NLL of the sequence at the current iteration, assessed by the ESM-2-650M protein language model. The shaded area around the traced trajectory represents the standard deviation of the NLL at t
Figure 2: Probability paths taken by models under various sampling budgets $(N)$ . The y-axis represents the NLL of the sequence at the current iteration, assessed by the ESM-2-650M protein language model. The shaded area around the traced trajectory represents the standard deviation of the NLL at t

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。