HIV Protease Database (HIVdb) is a repository for those structures of HIV protease that have never been released or deposited to the Protein Data Bank (PDB). Together with the official PDB data, HIVdb provided a unique source of information in a statistical sense. The database contains 207 structures; 148 taken from PDB, and 59 that are unique entries in HIVdb. Query tools in terms of the creation of ensembles for statistical analysis were designed. We present a new form, location, tools and data form of the HIV Protease Database. The new tools utilize a standard PDB user interface, but provide extra capabilities connected exclusively with this one protein and its ligands. We also present a design strategy for a specific subset or sub-database of the PDB with the aim of pointing out the statistical dimension of the problem related to a single protein structure. We are currently annotating the ligands in order to include their chemical properties. This approach emphasises large scale databases and scalability.