用于蛋白质设计的 AlphaFold 蒸馏
AlphaFold Distillation for Protein Design
AI总结：
本文提出利用知识蒸馏将 AlphaFold 等正向折叠模型的置信度指标转化为端到端可微的蒸馏模型，作为结构一致性正则化器用于逆向蛋白质折叠训练，显著提升了序列恢复率和蛋白质多样性。
摘要：
逆向蛋白质折叠，即设计能够折叠成特定 3D 结构的序列的过程，在生物工程和药物发现中至关重要。传统方法依赖于实验解析的结构，但这些结构仅覆盖了蛋白质序列的一小部分。像 AlphaFold 这样的正向折叠模型通过从序列准确预测结构，提供了一种潜在的解决方案。然而，这些模型太慢，无法在训练期间集成到逆向折叠模型的优化循环中。为了解决这个问题，我们提出对折叠模型置信度指标（如 pTM 或 pLDDT 分数）使用知识蒸馏，以创建一个更快的、端到端可微的蒸馏模型。然后，该模型可以在训练逆向折叠模型时用作结构一致性正则化器。我们的技术用途广泛，可以应用于其他设计任务，例如基于序列的蛋白质填充。实验结果表明，我们的方法优于非正则化基线，在保持生成序列的结构一致性的同时，序列恢复率提高了 3%，蛋白质多样性提高了 45%。代码可在 https://github.com/IBM/AFDistill 获取。
英文摘要：
Inverse protein folding, the process of designing sequences that fold into a specific 3D structure, is crucial in bio-engineering and drug discovery. Traditional methods rely on experimentally resolved structures, but these cover only a small fraction of protein sequences. Forward folding models like AlphaFold offer a potential solution by accurately predicting structures from sequences. However, these models are too slow for integration into the optimization loop of inverse folding models during training. To address this, we propose using knowledge distillation on folding model confidence metrics, such as pTM or pLDDT scores, to create a faster and end-to-end differentiable distilled model. This model can then be used as a structure consistency regularizer in training the inverse folding model. Our technique is versatile and can be applied to other design tasks, such as sequence-based protein infilling. Experimental results show that our method outperforms non-regularized baselines, yielding up to 3% improvement in sequence recovery and up to 45% improvement in protein diversity while maintaining structural consistency in generated sequences. Code is available at https://github.com/IBM/AFDistill
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
