Two classes of graph-theoretic molecular descriptors, viz., topological indices (TIs) and atom pairs APs), have been used to derive high-quality quantitative structure-activity relationships (QSARs) for inhibitors of dihydrofolate reductases (DHFRs) isolated from the wild and four mutant strains of Plasmodium falciparum. Of the three methods used for QSAR formulation, viz., principal-components regression (PCR), partial least squares (PLS), and ridge regression (RR), the RR method outperformed the other two. Cohen's kappa values, based on the overlap of significant and insignificant structural descriptors calculated for the QSAR development, show that DHFR from the wild strain is substantially different from the four mutant strains. The differential QSAR approach reported in this study can be used in protocols for the development of drugs to combat drug-resistant pathogens arising continuously in nature due to mutations. The pairwise kappa values in conjunction with appropriate drug targets and their corresponding set of ligands may be a useful tool in gauging the evolving mutual similarities and dissimilarities of pathogenic organisms from purely computed mathematical descriptors of the ligands.