Single-cell RNA sequencing data analysis using Explainable ArtificialIntelligence identified key transcriptional factors for early COVID-۱۹severity prediction

Early and precise diagnosis of SARS-CoV-۲ infection can be critical for managing the COVID-۱۹ pandemic. But oneof the challenges with this disease’s administration is the lack of standardized methods for diagnosis and the incapa bility to guesstimate prognosis based on clinical features. Explainable Artificial Intelligence (XAI) has been one ofthe most widely discussed topics in recent years. It is widely used to tackle various issues across multiple domains inthe medical studies. Here, we have applied SHAP algorithm on a single cell RNA-Seq data to find hub genes in dif ferent stages of COVID-۱۹ pathophysiology. Firstly, a single-cell RNA-seq dataset had downloaded under the GEOnumber of GSE۱۶۵۰۸۰. To evaluate data, the Scanpy pipeline was employed. This data contains the samples of nor mal, severe, asymptomatic, severe recovery and moderate COVID-۱۹ patients. Single-cell RNA data processing wasperformed on cells that met the following criteria: (I) Each cell must contain at least ۵۰۰ genes and no more than۶۰۰۰۰ gene expression counts. (II) Less than ۲۰% of the genes counted must be mitochondrial genes. The scran pack age on Bioconductor was used to calculate normalized expression. At the next step, SHAP algorithm was used to findmost hub genes based on SHAP value in each status. Finally, the hub genes were imported in MsigDB to find theirenrichment and DGiDB to construct their related drugs. Our study revealed some transcriptional signature mea surable in blood samples, which discriminated between healthy people and COVID-۱۹ positive patients and showedpredictive value for later severity of COVID-۱۹ symptoms. This type of approaches could, by employing standardhospital laboratory analyses of patient blood, be utilized to identify, COVID-۱۹ patients at high risk of mortality

کلیدواژه ها:

Explainable Artificial Intelligence ، Single-Cell RNA-sequencing ، COVID-۱۹ ، SHAP