Sir

Analyses of the crystal structures of macromolecules, the coordinates of which are deposited in databases, constitute a growing component of the biochemical literature. Given the magnitude of investment that has been made in determining protein and nucleic-acid structures, this is entirely appropriate; the gold hidden in that vast mine of information needs to be recovered. But I am disturbed by a practice of the community that engages in coordinate analysis: its use of database identification numbers as the sole reference to the coordinate sets discussed in publications.

This is not good scholarly practice. It conceals from readers the identities of the scientists responsible for the coordinate sets used, and makes it difficult for readers to find primary references. It is also unfair. For better or worse, citation indices are increasingly used to evaluate the contributions scientists have made to their fields. No credit will accrue to those who made the effort to determine a structure unless the papers that make use of its coordinates include a proper reference.

For these reasons, the Biophysical Journal will from now on require authors to include in their papers full references for all the coordinate data sets they have used, as well as database identification numbers. I hope that other journals will institute similar policies to keep what is now a modest problem from getting out of hand.