Target Fishing and Molecular Docking of Terpenes: A Systematic Evaluation of Method Concordance and Ligand Efficiency

Abstract Terpenes are naturally occurring small molecules with diverse chemical structures and potential therapeutic applications. Despite their wide-ranging reported bioactivities including antimicrobial, anti-inflammatory, analgesic, and anticancer properties, the molecular targets underlying these effects remain poorly characterized for most terpenes. Here, we apply a combined in silico pipeline, ligand-based target fishing using four complementary tools, followed by molecular docking, to 23 terpenes spanning monoterpene and sesquiterpene classes, with the dual goals of prioritizing candidate protein targets and benchmarking the agreement between the two computational methods. Consensus target fishing across four tools identified five relatively high-frequency protein targets: cannabinoid receptor 2 (CB2), androgen receptor (AR), 11β-hydroxysteroid dehydrogenase 1 (11β-HSD1), cytochrome P450 19A1 (CYP19A1), and estrogen receptor β (ERβ). Molecular docking against these targets revealed that known high-affinity ligands consistently outscored terpenes for four of five targets, while several terpenes exceeded known-ligand scores for CYP19A1. Critically, the correspondence between target fishing predictions and docking performance was inconsistent: target-fishing-positive terpenes did not reliably rank higher in docking than target-fishing-negative ones, and enrichment analysis confirmed that docking scores carried limited discriminatory power in this chemical series. These findings describe both the promise and the current limitations of integrated in silico approaches for terpene pharmacology and provide a ranked set of experimentally testable terpene-target interactions. We conclude that it is imperative to develop, curate, and distribute more data on terpenes to enable the pharmacological development of that rich class of natural compounds.