Megalodon模型通过模块化协同设计，显著提升3D分子生成质量与能量合理性。
Applications of Modular Co-Design for De Novo 3D Molecule Generation
- 基于等变层与联合连续离散去噪目标，构建可扩展的Transformer架构
- 参数增至40M后，有效大分子生成量提升49倍，能量降低2-10倍
- 在3D结构真实性和能量评估上达到当前最优，适合药物设计场景
从头生成3D分子是药物发现中的关键任务。然而，许多现有几何生成模型虽保持2D结构合法与拓扑稳定，却难以生成高质量3D构型。为此，我们提出Megalodon——一系列可扩展的Transformer模型，融合基础等变层，并采用联合连续与离散去噪的协同设计目标进行训练。我们在已有分子生成基准上评估其性能，并引入新的3D结构基准，重点考察模型生成真实分子结构的能力，特别是能量特性。结果表明，Megalodon在3D分子生成、条件结构生成及结构能量基准上均达当前最优水平，适用于扩散与流匹配方法。当参数量增至40M时，生成有效大分子数量提升49倍，能量水平较最佳前序模型低2-10倍。
原文摘要 · Abstract (English)
De novo 3D molecule generation is a pivotal task in drug discovery. However, many recent geometric generative models struggle to produce high-quality 3D structures, even if they maintain 2D validity and topological stability. To tackle this issue and enhance the learning of effective molecular generation dynamics, we present Megalodon-a family of scalable transformer models. These models are enhanced with basic equivariant layers and trained using a joint continuous and discrete denoising co-design objective. We assess Megalodon's performance on established molecule generation benchmarks and introduce new 3D structure benchmarks that evaluate a model's capability to generate realistic molecular structures, particularly focusing on energetics. We show that Megalodon achieves state-of-the-art results in 3D molecule generation, conditional structure generation, and structure energy benchmarks using diffusion and flow matching. Furthermore, doubling the number of parameters in Megalodon to 40M significantly enhances its performance, generating up to 49x more valid large molecules and achieving energy levels that are 2-10x lower than those of the best prior generative models.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。