用分子指纹预训练提升图神经网络在药物活性预测中的表现
On Improving Graph Neural Networks for QSAR by Pre-training on Extended-Connectivity Fingerprints
- 用扩展连接指纹（ECFP）预训练图神经网络，增强分子表征能力
- 在五项生物技术公司基准测试中显著优于现有方法，尤其在分布外数据上
- 揭示子结构泄漏风险，但整体仍适用于多种真实药物研发任务
分子图神经网络（GNN）在药物发现和定量构效关系（QSAR）研究中日益普及，但其相比传统分子特征化方法的优势仍存争议。本文提出一种通用策略：通过预训练使GNN预测扩展连接指纹（ECFP），以提升其在QSAR任务中的性能。在五个生物技术公司（Biogen）基准数据集上，采用ECFP预训练的GNN在标准评估指标上显著优于所有基线模型，且统计检验显著。然而，在更异质的数据集及复杂终点（如结合亲和力预测）的分布外（OOD）设置下，预训练模型表现较差。我们进一步分析了预训练阶段子结构级数据泄露对下游性能的影响。尽管存在某些场景下预训练效果受限，结果表明基于ECFP的预训练仍可有效提升多样化的实际QSAR任务在分布外情况下的表现。
原文摘要 · Abstract (English)
Molecular Graph Neural Networks (GNNs) are increasingly common in drug discovery, particularly for Quantitative Structure-Activity Relationship (QSAR) studies; yet, their superiority compared to classical molecular featurisation approaches is disputed. We report a general strategy for improving GNNs for QSAR by pre-training to predict Extended-Connectivity Fingerprints (ECFP). We validate our approach with statistical tests and challenging out-of-distribution (OOD) splits. Across five out of six Biogen benchmarks, we observed a statistically significant improvement in standard performance metrics over all evaluated baselines when using ECFP pre-trained GNNs. However, for more heterogeneous datasets and more complex endpoints, such as binding affinity prediction, pre-trained GNNs underperformed in OOD settings. Importantly, we investigated the impact of substructure-level data leakage during pre-training on downstream performance. While we identified scenarios where pre-training on ECFPs was less effective, our findings show that ECFP-based pre-training can enhance downstream OOD performance on a diverse set of practically relevant QSAR tasks.
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。