Structure-based drug discovery relies on efficient compilation and analysis of chemical, biological, pre-clinical and structural data. Factors such as inconsistent naming standards, publications and private archives with scattered and incomplete data, Web pages with little or no cross-talks act against this drug-design effort. Large scale effort on AIDS drug discovery has led to many issues ranging from archival and distribution of the information to annotation and naming standards for the data. To alleviate this situation we have been working on developing Web based public data resources (HIV Structural Database (http://bioinfo.nist.gov/SemanticWeb_pr2d/chemblast.do
). A significant part of the data found in these data resources is specific to the resources and they were obtained by examining literature or by direct request to individual researchers. To our knowledge, these Web resources have the largest collection of pre-clinical data and this effort may serve as a model for future efforts focused on other diseases.