Exogenous Chemicals Impact Virus Receptor Gene Transcription: Insights from Deep Learning

Despite the fact that coronavirus disease 2019 (COVID-19),causedby severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), hasbeen disrupting human life and health worldwide since the outbreakin late 2019, the impact of exogenous substance exposure on the viralinfection remains unclear. It is well-known that, during viral infection,organism receptors play a significant role in mediating the entryof viruses to enter host cells. A major receptor of SARS-CoV-2 isthe angiotensin-converting enzyme 2 (ACE2). This study proposes adeep learning model based on the graph convolutional network (GCN)that enables, for the first time, the prediction of exogenous substancesthat affect the transcriptional expression of the ACE2 gene. It outperformsother machine learning models, achieving an area under receiver operatingcharacteristic curve (AUROC) of 0.712 and 0.703 on the validationand internal test set, respectively. In addition, quantitative polymerasechain reaction (qPCR) experiments provided additionalsupporting evidence for indoor air pollutants identified by the GCNmodel. More broadly, the proposed methodology can be applied to predictthe effect of environmental chemicals on the gene transcription ofother virus receptors as well. In contrast to typical deep learningmodels that are of black box nature, we further highlight the interpretabilityof the proposed GCN model and how it facilitates deeper understandingof gene change at the structural level. A novel deeplearning architecture effectively evaluatesimpact of exogenous chemicals on virus receptor gene expression andwith certain interpretability.