Currently, DARTS reflects the RNA structural repertoire of the 1333 RNA solved structures available in the March 2007 release of the PDB. DARTS provides a non-redundant dataset of 244 structures that represents the whole data (maximum 90% structural or sequential identity). Each of the remaining structures is assigned to one representative structure of the non-redundant dataset as either a homolog or a homologous fragment (the difference between a homolog or a homologous fragment is the size ratio between the two structures).

Additionally, DARTS provides a hierarchical classification of all the 1333 clusters into 94 clusters. The description of each cluster is a dendrogram of its non-redundant members (representaives) and the representaives' homologs and homologous fragments.

The hierarchical clustering method used to construct DARTS is fully automated and mainly on the basis of their spatial similarity determined by ARTS.