Lipid nanoparticles (LNPs) formulated with ionizable lipids, such as ALC-0315 and SM-102, suffer from significant hepatic off-target expression following intramuscular administration, which could potentially impair hepatic function and reduce therapeutic efficacy. Currently, novel ionizable lipid discovery predominantly relies on stochastic high-throughput screening. Here, we designed a 2-aminoimidazole (AM)-based ionizable lipid library and established a molecular dynamics-guided strategy for quantitatively analyzing the lipid-mRNA interaction strength, which might offer a rational screening alternative. A33-D268 emerged as the top-performing candidate through molecular dynamics simulations. Following formulation optimization via definitive screening design experiments, the AM-LNP/mRNA formulated with A33-D268 demonstrated an intramuscular transfection efficiency similar to that of ALC-0315-based LNPs, with markedly reduced hepatic off-target expression.