A novel concept titled ‘Peptide Mold’ for mapping potential binding sites in protein targets is presented. A large multiconformer tetrapeptide library comprising of 32 million conformations of all possible combinations of naturally-occurring amino acids was constructed and used for molecular docking analysis in the substrate-binding site of SARS-CoV-2 PLpro enzyme. The top-ranking, structurally-diverse tetrapeptide docked conformations (symbolizing peptide mold, analogous to a clay mold) were used then for elucidating a five-point pharmacophore. Ligand-based virtual screening of a large, multiconformer library of phytoconstituents using the derived five-point pharmacophore led to identification of potential binders for SARS-CoV-2 PLpro at its substrate-binding site. The approach is based on generating the imprint of a macromolecular binding site (cavity) using tetrapeptides (clay), thereby generating a reverse mold (with definitive shape and size), which can further be used for identifying small-molecule ligands matching the captured features of the target binding site. The approach is based on the fact that the individual amino acids in the tetrapeptide represent all possible drug-receptor interaction features (electrostatic, H-bonding, van der Waals, dispersion and hydrophobic among others). The ‘peptide mold’ approach can be extended to any protein target for mapping the binding site(s), and further use of the generated pharmacophore model for virtual screening of potential binders. The peptide mold approach is a robust, hybrid computational screening strategy, overcoming the present limitations of structure-based methods, e.g., molecular docking and the ligand-based methods such as pharmacophore search. Exploration of the peptide mold strategy is expected to yield high-quality, reliable and interesting virtual hits in the computational screening campaigns during the hit and lead identification stages.