For AIDS therapy, there are currently a number of compounds available for multiple targets already approved by the FDA and in clinic, e.g. protease inhibitors, reverse transcriptase inhibitors (NRTI, NNRTI), fusion inhibitors, CCR4, CCR5 among others. Some pharmaceuticals act against the virus before the entrance of HIV into the host cells. One of these targets is the glucosidase protein. This novel fusion target has been recently explored because the synthesis of viral glycoproteins depends on the activity of enzymes, such as glucosidase and transferase, for the elaboration of the polysaccharides. In this work we have built an homology model of Saccharomyces cerevisiae glucosidase and superimposed all relevant glucosidase-like enzymes in complex with carbohydrates, and calculated as well molecular interaction fields in our S. cerevisiae active site model. Our results suggest that there are two saccharide binding sites which are the most important for the binding of inhibitors with this family of enzymes which supports the possibility of inhibitors containing only two sugar units. Based on these results, we have proposed a novel pseudo-dissacharide which is a potential pharmaceutical for AIDS treatment.