论文标题
基于动力学的肽-MHC结合优化通过卷积变异自动编码器：Castelo的用例模型
Dynamics-based peptide-MHC binding optimization by a convolutional variational autoencoder: a use-case model for CASTELO
论文作者
论文摘要
抗原特异性免疫疗法发展中未解决的挑战是确定靶标的最佳抗原。理解抗原-MHC结合对于实现这一目标至关重要。在这里，我们提出了Castelo，这是一种合并的机器学习 - 分子动力学（ML-MD）方法，用于设计1型糖尿病（T1D）刺激系统的MHC结合亲和力增加的新型抗原。我们通过在48种抗原和4种HLA血清型的48个不同系统的MD轨迹上训练卷积变化自动编码器（CVAE），以小分子铅优化算法为基础。我们开发了几个新的机器学习指标，包括基于结构的锚固残基分类模型以及群集比较分数。 ML-MD预测与实验结合结果和自由能扰动预测的结合亲和力非常吻合。此外，ML-MD指标独立于传统的MD稳定性指标，例如接触区域和RMSF，这些指标不反映结合亲和力数据。我们的工作支持基于结构的深度学习技术在抗原特异性免疫疗法设计中的作用。
An unsolved challenge in the development of antigen specific immunotherapies is determining the optimal antigens to target. Comprehension of antigen-MHC binding is paramount towards achieving this goal. Here, we present CASTELO, a combined machine learning-molecular dynamics (ML-MD) approach to design novel antigens of increased MHC binding affinity for a Type 1 diabetes (T1D)-implicated system. We build upon a small molecule lead optimization algorithm by training a convolutional variational autoencoder (CVAE) on MD trajectories of 48 different systems across 4 antigens and 4 HLA serotypes. We develop several new machine learning metrics including a structure-based anchor residue classification model as well as cluster comparison scores. ML-MD predictions agree well with experimental binding results and free energy perturbation-predicted binding affinities. Moreover, ML-MD metrics are independent of traditional MD stability metrics such as contact area and RMSF, which do not reflect binding affinity data. Our work supports the role of structure-based deep learning techniques in antigen specific immunotherapy design.
