Based on genome wide association studies (GWAS), the activities of phosphodiesterase 4D (PDE4D) and 5-Lipoxygenase activating protein (ALOX5AP) were suggested as two of the major factors involved in ischemic stroke risks. Uncontrolled PDE4D activities often lead to cAMP-induced stroke and cardiovascular diseases. Overexpression of ALOX5AP, on the other hand, had been shown to play a major role in inflammation pathway that could induce the development of atherosclerosis and stroke. To eliminate the risk factors that lead to stroke, we reported the identification and analysis of dual-targeting compounds that could reduce PDE4D and ALOX5AP activities from traditional Chinese medicine (TCM). We employed world's largest TCM database, TCM Database@Taiwan, for in silico drug identification. We also introduced machine learning predictive models, as well as pharmacophore model, for characterizing the drug-like candidates. Both myristic acid and pentadecanoic acid were identified. The follow-up analysis on molecular dynamics simulation further determined the major roles of the carboxyl group for forming stable molecular interactions. Intriguingly, the carboxyl group demonstrated different bonding patterns with PDE4D and ALOX5AP, through electrostatic interaction and hydrogen bonds, respectively. In addition, the large volume occupied by the ligand hydrophobic regions could achieve inhibition through occupying the vacant spaces in the binding site. These pharmacophores held true for both candidates against each protein targets. Hence, we proposed the presence of the carboxyl group and hydrophobic regions as potent dual targeting features that inhibit both PDE4D and ALOX5AP activities.