The aim of this work was to optimize, by means of molecular modeling software, biomimetic-based traps for pathogen detection suitable for analytical applications like screening or pre-analytical methods. The pathogen prototype system chosen was Listeria monocytogenes because of the large number of X ray and NMR structures available. 298 oligopeptides were computationally designed mimicking the binding pocket of the mammalian protein E-cadherin, the target of Listeria monocytogenes adhesion, internalin A. The contribution of individual peptides to bind was investigated using FRED, a protein-ligand docking program. Ten peptides were selected for experimental analysis taking as selection parameters the length, the position in the docking pocket and the score of simulated binding energy. A series of competition assays were carried out using high density colorimetric microarray using various bacteria species (Listeria monocytogenes, Listeria monocytogenes genetically modified without internalin A, Listeria innocua and Lactococcus lactis) in solution with computationally selected peptides. The data demonstrated that peptides could be able to distinguish Listeria monocytogenes with an EC50 up to 107cfu × mL−1. In particular the peptide with the best calculated binding score gave the highest statistically unambiguous response toward Listeria monocytogenes compared to other bacteria, demonstrating that rationally simulated approach can be useful as preliminary screening in the choice of biomimetic ligands.