Hi Rachel, This may be just a personal opinion (others, feel free to pipe up!), but I would go for representatives as opposed to averages. Presumably each ensemble member has reasonable internal coordinates, whereas an average may be distorted. The average conformation may never be sampled in reality even if each of the conformations contributing to the average do exist. Thus, I like your last option the best. However, factors to consider are the dispersion of the ensemble, the averaging method, whether the average has been minimized, and how similar the representative is to the average. You may also wish to consult the literature and see what others have done in similar situations. Best, Elaine ----- Elaine C. Meng, Ph.D. meng@cgl.ucsf.edu UCSF Computer Graphics Lab and Babbitt Lab Department of Pharmaceutical Chemistry University of California, San Francisco http://www.cgl.ucsf.edu/home/meng/index.html On Jan 29, 2008, at 12:24 PM, Rachel Scherer wrote:

Given that chimera doesn't calculate average NMR structures, this query may be outside the groups interests, but can anyone suggest what the appropriate protocol is when making aa-substitutions into an NMR structure: should an "average" be calculated (by averaging the atom coordinates across all the models in a PDB file that predates inclusion of the average into the database entry) or, should one use a representative from a cluster generated with "ensemble cluster" be chosen? Or an average of the representatives of the clusters? Or, the cluster representative that best matches the calculated average? --

Rachel Scherer rachel@bio.umass.edu Biology Department (413)545-2825 vox UMASS Amherst (413)545-3243 fax