The stochastic tunneling technique is applied to screen a database of chemical compounds to the active site of dihydrofolate reductase for lead candidates in the receptor-ligand docking problem. Using an atomistic force field we consider the ligand's internal rotational degrees of freedom. It is shown that the natural ligand (methotrexate) scores best among 10000 randomly chosen compounds. We analyze the top scoring compounds to identify hot-spots of the receptor. We mutate the amino acids that are responsible for the hot-spots of the receptor and verify that its specificity is lost upon modification. (C) 2003 Elsevier Science B.V. All rights reserved.