RELIABILITY OF PROTEIN STRUCTURE DETERMINATION. T. Alwyn Jones & Gerard J. Kleywegt. Department of Molecular Biology, Uppsala University, Biomedical Centre, Box 590, S-751 24 Uppsala, Sweden

The reliability of protein structure determination depends on many factors, not least of which is the experience of the crystallographers involved. The kinds of errors introduced into a model during the initial map interpretation will be discussed. The identification of these errors requires a refinement protocol that is suited to each study. Different protocols depend primarily on the resolution of the study, and the number of NCS copies in the asymmetric unit. A policy of constant model evaluation during the refinement process is needed in all protocols. We suggest that the community can do a better job of refining macromolecular structures. Whether the community actually does a better job, seems to depend on the policy controlling the deposition of structure factor data.