Integrative transcriptome analysis of SARS‐CoV‐2 human‐infected cells combined with deep learning algorithms identifes two potential cellular targets for the treatment of coronavirus disease.

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) quickly spread worldwide, leading coronavirus disease 2019 (COVID-19) to hit pandemic level less than 4 months after the frst ofcial cases. Hence, the search for drugs and vaccines that could prevent or treat infections by SARS-CoV-2 began, intending to reduce a possible collapse of health systems. After 2 years, eforts to fnd therapies to treat COVID-19 continue. However, there is still much to be understood

about the virus’ pathology. Tools such as transcriptomics have been used to understand the impact of SARS-CoV-2 on dif- ferent cells isolated from various tissues, leaving datasets in the databases that integrate genes and diferentially expressed

pathways during SARS-CoV-2 infection. After retrieving transcriptome datasets from diferent human cells infected with SARS-CoV-2 available in the database, we performed an integrative analysis associated with deep learning algorithms to determine diferentially expressed targets mainly after infection. The targets found represented a fructose transporter (GLUT5) and a component of proteasome 26s. These targets were then molecularly modeled, followed by molecular docking that identifed potential inhibitors for both structures. Once the inhibition of structures that have the expression increased by the virus can represent a strategy for reducing the viral replication by selecting infected cells, associating these bioinformatics tools, therefore, can be helpful in the screening of molecules being tested for new uses, saving fnancial resources, time, and making a personalized screening for each infectious disease.