Identification of a novel drug target in Porphyromonas gingivalis by a computational genome analysis approach

Abstract

Abstract This study applied a subtractive genomics approach to identify a potential drug target in the Porphyromonas gingivalis strain (ATCC BAA-308/W83). The aim was to characterize the whole proteome and hypothetical proteins (HPs) through structural, functional, and pathway predictions. The proteome was systematically reduced to identify essential proteins (EPs), non-homologous proteins (NHPs), and non-paralogous proteins (NPPs) while excluding those that were similar to the human proteome. Out of 1,836 proteins, the cluster database at high identity with tolerance algorithm identified 36 sequences as paralogous, having 80% identity. The resulting 1,827 proteins were compared to the human proteome using BLASTp (e-value 10−3), resulting in 1,427 NHPs. These were then aligned with the DEG database using BLASTp (e-value of 10−5), identifying 396 NHPs essential for pathogen survival. CELLO predicted the sub-cellular localization, and KEGG Automated Annotation Server identified potential metabolic pathways using a BLASTp similarity search of NHPs and EPs against the infrequently updated KEGG database. A total of 79 HPs essential for P. gingivalis were selected, and their molecular weights were determined. HPs were screened for metabolic pathway prediction, and the 3D structures of the proposed HPs were determined using homology modeling, and validation was performed. Only one HP (putative arginine deiminase) was qualified and found to be involved in the arginine and proline metabolic pathway.