Potential drug target from a hypothetical protein of African swine fever virus: an in silico analysis

Abstract

African swine fever virus remains a threat to pig farmers all over the world, and there are currently no treatments for African swine fever. In silico analysis of a hypothetical protein of the African swine fever virus was conducted to identify potential drug ligands. Initial drug ligand screening for potential protein identities was performed with DrugBank sequence search. Physicochemical properties of the target hypothetical protein were predicted with ProtParam. Conserved domains were predicted with CDD-BLAST, and a phylogenetic tree was constructed with MEGA X. A 3D homology model was created with SWISS-MODEL, the structural validity of which was checked with PROCHECK, ProSA-web, and QMEAN. Location of the protein within the host cell and protein topology were predicted with Virus-mPloc and Phobius, respectively. Binding sites were predicted with Residue DEPTH and 3DLigandSite. Drug docking was performed with MTIAutoDock and additional ligand screening was performed with MTIOpenScreen. Results showed that the hypothetical protein belonged to the serine/threonine-protein kinase family and contained conserved domains found in the human PIM1 kinase protein, which also served as the template for homology modelling. Literature suggests the possible role of the hypothetical protein in viral replication. Virtual screening identified intermediate drugs, and ligand docking showed that quercetin, and fostamatinib, are the optimal target drugs.