Organophosphate poisoning can occur from exposure to agricultural pesticides or chemical weapons. This exposure inhibits acetylcholinesterase resulting in increased acetylcholine levels within the synaptic cleft causing loss of muscle control, seizures, and death. Mitigating the effects of organophosphates in our bodies is critical and yet an unsolved challenge. Here, we present a computational strategy that integrates structure mining and modeling approaches, using which we identify novel candidates capable of interacting with a serine hydrolase probe (with equilibrium binding constants ranging from 4 to 120 μM). One candidate Smu. 1393c catalyzes the hydrolysis of the organophosphate omethoate (kcat/Km of (2.0 ± 1.3) × 10-1 M-1s-1) and paraoxon (kcat/Km of (4.6 ± 0.8) × 103 M-1s-1), V- and G-agent analogs respectively. In addition, Smu. 1393c protects acetylcholinesterase activity from being inhibited by two organophosphate simulants. We demonstrate that the utilized approach is an efficient and highly-extendable framework for the development of prophylactic therapeutics against organophosphate poisoning and other important targets. Our findings further suggest currently unknown molecular evolutionary rules governing natural diversity of the protein universe, which make it capable of recognizing previously unseen ligands.