Designing a Prognostics Framework for Pharmaceutical Development: Applying PHM Principles to Computational Drug Discovery for Novel HIV-1 C(SA) Protease Inhibitors

Abstract

Prognostics and Health Management (PHM) has transformed the engineering industry through accurate pre-initialisation failure prediction. Here, those principles are applied to pharmaceutical development by treating a physics-based computational chemistry model as a pre-synthesis failure predictor for candidate drug molecules. The target is the HIV-1 subtype C (South African, C(SA)) protease, the dominant strain across sub-Saharan Africa, against which the existing subtype-B-optimised protease inhibitors lose potency. Twenty pentacycloundecane (PCU) cage peptoid candidates in the C→N backbone orientation were evaluated against the HIV-1 C(SA) protease (PDB 3U71) by molecular dynamics and MM-GBSA binding free energy calculations (AMBER 20, ff19SB, n = 4 replicates), with nine FDA-approved protease inhibitors evaluated under an identical protocol as positive controls. Glu-PCU-Glu ranked first (ΔGbind = −89.24 ± 7.42 kcal mol⁻¹); the conformationally constrained Pro-PCU-Pro ranked last (−22.22 ± 2.20 kcal mol⁻¹). The MM-GBSA screen functions as a Stage 1 health check that rules out the primary failure mode, insufficient target affinity, before any synthesis cost is incurred. Three compounds, Glu-PCU-Glu, Cys-PCU-Cys and Tyr-PCU-Tyr, pass the check and are carried forward.

Keywords:  prognostics and health management; pre-initialisation failure prediction; computational drug discovery; MM-GBSA; HIV-1 C(SA) protease