基于图神经网络协同过滤的药物疾病关联预测

Collaborative filtering based on graph neural network for drug-disease association prediction

新药开发需要耗费很高的成本,建立高效且高准确度的药物新适应症预测方法非常重要,提出一种基于图神经网络协同过滤的药物疾病关联预测方法,获取药物与疾病治疗关系中的信息并结合药物相似性获得更好的预测表现。首先通过图神经网络提取药物-疾病治疗关系数据中的协作信号细化药物嵌入,然后利用嵌入计算药物之间的治疗关系相似性,再结合药物化学结构、蛋白质和副作用相似性预测药物的新作用。与现有的协同过滤方法在相同数据集上进行对比,本文方法获得了较高的预测精确率(0.664 8)。所提出的获取药物-疾病治疗关系中的潜在信息并结合相似性进行药物疾病关联预测的策略是有效的,有助于发现药物的新适应症,为药物开发提供帮助。

Development of new drugs takes a long time and is high-cost.Hence it's critical to have efficient and precise methods for predicting new indications for drugs.A drug-disease association prediction method based on graph neural collaborative filtering is proposed in an attempt to obtain information in drug-disease treatment relationships and combine with drug similarities for obtaining better prediction performance.The proposed method firstly capture collaborative signals in drug-disease treatment relationships through graph neural network to refine drug embeddings,then use the drug embeddings to calculate the similarities in drug-disease treatment relationships between drugs,and finally combines with the similarities in drug chemical structures,proteins,and side effects to predict drug repurposing.Compared with the existing collaborative filtering methods on the same data set,the proposed method achieves a higher prediction accuracy(0.664 8).The proposed strategy to obtain potential information in drug-disease treatment relationships and combine with similarities for drug-disease association prediction is effective and helps to discover new indications for drugs and provides assistance in drug development.