Network-based bioinformatics for the prediction of candidate SNPs in Staphylococcus aureus virulence and resistance...

This study proposes a network-based bioinformatics strategy to predict candidate genes for SNPs in Staphylococcus aureus that act as points of convergence between antimicrobial resistance and biofilm formation. Based on 19 Staphylococcus aureus resistance and virulence genes, a protein-protein interaction network was constructed on the STRING platform and expanded with 20 first-degree interactors. The topology revealed central hubs with high connectivity, such as IcaA (degree 18), Atl (14), SarA (11), ClfA/FnbA (15-21), and low-degree proteins such as vraSR (2) and mgrA (4). MCL analysis divided the network into 10 functional clusters; Cluster 1 grouped adhesion and biofilm proteins together with the mecA resistance gene, highlighting molecular integration. Functional enrichment (Gene Ontology) showed significant over-representation of cell adhesion (FDR = 1.83 × 10⁻⁶) and transcriptional regulation (FDR = 0.042), with an overall interaction p-value < 1.0 × 10⁻¹⁶. The hubs were stratified into regulators (Group 1: SarA, MgrA, VraS), with a cascade effect, and effectors (Group 2: IcaA, Atl, ClfA/FnbA), with a direct effect on the structure. These six predicted candidate genes may represent the main points of convergence between resistance mechanisms and biofilm formation, constituting priority targets for genomic association studies and for the development of new therapeutic strategies.