Underexplored ligand-binding features of FabI from S. aureus and E. coli: a comparative pharmacophoric modeling and surface mapping approach

All FabI structures were retrieved from the PDB (provided as an extended dataset in this Zenodo together with filtered metadata). Structures were treated according to (REF) and multimers were split into monomers and aligned by their backbone using Maestro (2024v3). All-against-all pairwise structure comparisons of FabI PDB structures were carried out using Template Modeling (TM)-score computed by the TM-align algorithm implemented in the US-align standalone with the monomeric structure alignment option. TM-align calculates two TM-scores normalized by the length of each of a pair of structures. The mean of the two TM-scores was calculated to represent the pair. This matrix of pairwise distance values was used on a multidimension scaling analysis (MDS).