We have been working on ways to bring to the forefront, important research. We have done this with regards to the general scientific literature through our myriad of life science portals. This is the first attempt to rank 3D structures according to their popularity among scientists and citation count. We have used 2 data sources to create our dataset. The first set of data has come from citation data extracted from the Pubmed Central database. In addition, we have used the dataset provided by the Worldwide Protein Data Bank which contains data regarding the number of downloads of each structure over time via their different web services.

Scoring of structures involves a simple equation in which the product of the total number of citation and the total number of downloads (scaled to not overwhelm the citation data) is divided by the squareroot of the number of days since the structure was released. The time-scaling ensures that older structures which have had time to accumulate more downloads and citations are not given an unfair advantage. The squareroot modifier ensures that newer structures are not given an unfair advantage due to the fact that citation and download rates have been increasing over time (data not shown).