School of Software, Yunnan University, Kunming 650091, China.
Bioinformatics. 2020 Apr 15;36(8):2538-2546. doi: 10.1093/bioinformatics/btz965.
Predicting the association between microRNAs (miRNAs) and diseases plays an import role in identifying human disease-related miRNAs. As identification of miRNA-disease associations via biological experiments is time-consuming and expensive, computational methods are currently used as effective complements to determine the potential associations between disease and miRNA.
We present a novel method of neural inductive matrix completion with graph convolutional network (NIMCGCN) for predicting miRNA-disease association. NIMCGCN first uses graph convolutional networks to learn miRNA and disease latent feature representations from the miRNA and disease similarity networks. Then, learned features were input into a novel neural inductive matrix completion (NIMC) model to generate an association matrix completion. The parameters of NIMCGCN were learned based on the known miRNA-disease association data in a supervised end-to-end way. We compared the proposed method with other state-of-the-art methods. The area under the receiver operating characteristic curve results showed that our method is significantly superior to existing methods. Furthermore, 50, 47 and 48 of the top 50 predicted miRNAs for three high-risk human diseases, namely, colon cancer, lymphoma and kidney cancer, were verified using experimental literature. Finally, 100% prediction accuracy was achieved when breast cancer was used as a case study to evaluate the ability of NIMCGCN for predicting a new disease without any known related miRNAs.
https://github.com/ljatynu/NIMCGCN/.
Supplementary data are available at Bioinformatics online.
预测 microRNA（miRNA）与疾病之间的关联在识别与人类疾病相关的 miRNA 方面起着重要作用。由于通过生物实验鉴定 miRNA-疾病关联既耗时又昂贵，因此目前使用计算方法作为确定疾病与 miRNA 之间潜在关联的有效补充。
我们提出了一种新的基于图卷积网络的神经归纳矩阵补全方法（NIMCGCN）来预测 miRNA-疾病关联。NIMCGCN 首先使用图卷积网络从 miRNA 和疾病相似性网络中学习 miRNA 和疾病的潜在特征表示。然后，将学习到的特征输入到一个新的神经归纳矩阵补全（NIMC）模型中，生成关联矩阵补全。NIMCGCN 的参数是基于已知的 miRNA-疾病关联数据进行端到端的监督学习得到的。我们将提出的方法与其他最先进的方法进行了比较。接收者操作特征曲线下的面积结果表明，我们的方法明显优于现有方法。此外，使用实验文献验证了三种高危人类疾病（结肠癌、淋巴瘤和肾癌）前 50 个预测 miRNA 中的 50、47 和 48 个。最后，当将乳腺癌作为案例研究来评估 NIMCGCN 预测没有任何已知相关 miRNA 的新疾病的能力时，实现了 100%的预测准确率。
https://github.com/ljatynu/NIMCGCN/。
补充数据可在生物信息学在线获得。
