当前位置: X-MOL 学术Nat. Mach. Intell. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A deep learning framework for high-throughput mechanism-driven phenotype compound screening and its application to COVID-19 drug repurposing
Nature Machine Intelligence ( IF 23.8 ) Pub Date : 2021-02-01 , DOI: 10.1038/s42256-020-00285-9
Thai-Hoang Pham 1 , Yue Qiu 2 , Jucheng Zeng 3 , Lei Xie 2, 4, 5, 6 , Ping Zhang 1, 3, 7
Affiliation  

Phenotype-based compound screening has advantages over target-based drug discovery, but is unscalable and lacks understanding of mechanism of drug action. A chemical-induced gene expression profile provides a mechanistic signature of phenotypic response; however, the use of such data is limited by their sparseness, unreliability and relatively low throughput. Few methods can perform phenotype-based de novo chemical compound screening. Here we propose a mechanism-driven neural network-based method, DeepCE—which utilizes a graph neural network and multihead attention mechanism to model chemical substructure–gene and gene–gene associations—for predicting the differential gene expression profile perturbed by de novo chemicals. Moreover, we propose a novel data augmentation method that extracts useful information from unreliable experiments in the L1000 dataset. The experimental results show that DeepCE achieves superior performances to state-of-the-art methods. The effectiveness of gene expression profiles generated from DeepCE is further supported by comparing them with observed data for downstream classification tasks. To demonstrate the value of DeepCE, we apply it to drug repurposing of COVID-19 and generate novel lead compounds consistent with clinical evidence. DeepCE thus provides a potentially powerful framework for robust predictive modelling by utilizing noisy omics data and screening novel chemicals for the modulation of a systemic response to disease.



中文翻译:

用于高通量机制驱动的表型化合物筛选的深度学习框架及其在 COVID-19 药物再利用中的应用

基于表型的化合物筛选比基于靶点的药物发现具有优势,但不可扩展且缺乏对药物作用机制的了解。化学诱导的基因表达谱提供了表型反应的机制特征;然而,此类数据的使用受到其稀疏性、不可靠性和相对较低的吞吐量的限制。很少有方法可以进行基于表型的从头化合物筛选。在这里,我们提出了一种机制驱动的基于神经网络的方法 DeepCE——它利用图神经网络和多头注意力机制来模拟化学子结构-基因和基因-基因关联——用于预测从头化学物质扰动的差异基因表达谱。而且,我们提出了一种新的数据增强方法,可以从 L1000 数据集中的不可靠实验中提取有用信息。实验结果表明,DeepCE 的性能优于最先进的方法。通过将 DeepCE 生成的基因表达谱与下游分类任务的观察数据进行比较,进一步支持了它们的有效性。为了证明 DeepCE 的价值,我们将其应用于 COVID-19 的药物再利用,并产生与临床证据一致的新型先导化合物。因此,DeepCE 通过利用嘈杂的组学数据和筛选用于调节对疾病的全身反应的新型化学物质,为稳健的预测建模提供了一个潜在的强大框架。通过将 DeepCE 生成的基因表达谱与下游分类任务的观察数据进行比较,进一步支持了它们的有效性。为了证明 DeepCE 的价值,我们将其应用于 COVID-19 的药物再利用,并产生与临床证据一致的新型先导化合物。因此,DeepCE 通过利用嘈杂的组学数据和筛选用于调节对疾病的全身反应的新型化学物质,为稳健的预测建模提供了一个潜在的强大框架。通过将 DeepCE 生成的基因表达谱与下游分类任务的观察数据进行比较,进一步支持了它们的有效性。为了证明 DeepCE 的价值,我们将其应用于 COVID-19 的药物再利用,并产生与临床证据一致的新型先导化合物。因此,DeepCE 通过利用嘈杂的组学数据和筛选用于调节对疾病的全身反应的新型化学物质,为稳健的预测建模提供了一个潜在的强大框架。

更新日期:2021-02-01
down
wechat
bug