Target breadth and mutation resilience in African-natural-product-inspired antimalarial chemotypes: a computational analysis

Abstract

Can target breadth and mutation resilience be evaluated as distinct, complementary properties when prioritizing antimalarial chemotypes from African natural-product-inspired chemical space? We addressed this question by linking chemical-space expansion to target-specific docking and a per-target resistance-resilience score (RRS), while keeping computational evidence separate from biological activity. A hybrid library of 65 856 molecules, generated from 396 African natural products and 454 synthetic antimalarial compounds, yielded 19 913 computationally prioritized synthesizable leads. A 17-member polypharmacology-oriented cohort was evaluated by AutoDock Vina against PfDHFR, PfCRT, PfClpP, and PfATP4 using target-specific structural anchors. All 68 candidate–target pairs passed the geometric pose-quality criterion, with scores ranging from −7.91 kcal/mol to −4.63 kcal/mol. Per-target RRS was calculated from separate PfDHFR (N51I, C59R, S108N, I164L) and PfCRT (K76T, K76A) mutant panels, excluding non-binding wild-type targets from the denominator. The cohort comprised six A*, five B, five C, and one D RRS profiles, spanning 68.2 to 111.7. Cross-metric associations were weak or non-significant after multiplicity correction: PNS–RRS ρ = −0.559, ACSI–RRS ρ = −0.132, and RRS–wild-type score ρ = −0.433. The resulting target profiles define testable hypotheses about chemotype breadth and mutation resilience; they do not establish potency, target engagement, or resistance circumvention.