Artificial intelligence-based imaging platform for COVID-19 infection of organoids

重新使用已批准用于人类的药物是一种有效的方法，可以识别具有既定药理学和安全性特征的SARS-CoV-2抑制剂，并可以快速转化为临床治疗COVID-19。基于人工智能（AI）的图像分析能够检测寨卡病毒感染细胞内质网的变化的先前工作，我们现在将应用基于AI的图像分析来检测由于SARS-CoV-2引起的内质网，自噬体和其他细胞器的改变，以便从再利用化合物库中识别抑制性候选药物。最初，肺上皮Calu-3细胞将在UBC生命科学研究所的3级（CL 3）FINDER设施中感染SARS-CoV-2，并使用AI识别与SARS-CoV-2感染相关的生物特征。然后，我们将筛选一个约3,000种抑制剂的再利用文库，以确定那些最能抑制SARS-CoV-2感染的抑制剂。随后，我们将通过我们的行业合作伙伴STEMCELL Technologies Inc.建立肾脏、大脑和肺气道类器官的二次筛选。以确定最好和最有前途的SARS-CoV-2抑制剂。我们将评估新的先导化合物与其他候选药物的协同作用。这笔赠款将支持开发最先进的人工智能图像分析，以加强SARS-CoV-2感染细胞和类器官的分类，并提供改进的定量方法来评估抗SARS-CoV-2药物的疗效。基于人工智能的图像分析方法的开发将支持开发新的治疗方法，以立即支持应对当前的COVID-19危机，并开发新的方法来成像多种不同的组织类器官，以应用于各种疾病模型。

Repurposing drugs already approved for use in humans is a valid approach to identify SARS-CoV-2 inhibitors with established pharmacological and safety profiles that can be rapidly translated to the clinic for treatment of COVID-19. Based on prior work in which artificial intelligence (AI)-based image analysis was able to detect changes to the endoplasmic reticulum of Zika virus infected cells, we will now apply AI-based image analysis to detect alterations to the endoplasmic reticulum, autophagosomes and other cellular organelles due to SARS-CoV-2 in order to identify inhibitory drug candidates from repurposing compound libraries. Initially, lung epithelial Calu-3 cells will be infected with SARS-CoV-2 in the containment level 3 (CL3) FINDER facility in the Life Sciences Institute at UBC and use AI to identify biosignatures associated with SARS-CoV-2 infection. We will then screen a repurposing library of ~3,000 inhibitors to identify those best able to inhibit SARS-CoV-2 infection. We will subsequently establish a secondary screen of kidney, brain and lung airway organoids, available through our industry partner STEMCELL Technologies Inc. to identify the best and most promising SARS-CoV-2 inhibitors. We will evaluate the new lead compounds for synergistic effect with other drug candidates. This grant will support the development of a state-of-the-art AI image analysis to enhance classification of SARS-CoV-2 infection of cells and organoids and provide improved, quantitative approaches to assess anti-SARS-CoV-2 drug efficacy. Development of AI-based image analysis approaches will support development of novel therapeutics for immediate support against the current COVID-19 crisis and also develop novel approaches to image multiple and varied tissue organoids for application to various disease models.