Hello Pan, """ I refined a cryo-EM data with 3.2 angstrom, the model fit well with the map, but the rotamer outlier is too high, about 8%.  How can I decrease the value of rotamer outlier by realspace.refinement in Phenix GUI, and no need to adjust the residues one by one in Coot? """ if you used latest version of Phenix, which is dev-4146 as of today: http://phenix-online.org/download/nightly_builds.cgi?show_all=1 then this is not expected, because phenix.real_space_refine has built-in ability to detect which residues need a fix and does an exhaustive grid-search in chi-torsion angle space to optimize it against the map and local environment. If this happened using the latest version of Phenix, then this is an indication to me that something has gone wrong and we need to investigate. For this I'm asking to send all input files. Residues that phenix.real_space_refine was unable to fix for whatever reason are best to fit manually in Coot. And I can't agree more with Tristan -- all the great tools and automation are to help you not substitute you! So, checking manually is important! """ Another question is the value of clash score are also too high. """ You did not tell us the value, so no answer here! A general guidance: the lower the better, but zero is not the goal. """ When I set the nonbonded_weight to 10000 """ According to study we conducted recently: http://phenix-online.org/presentations/nb_weight.pdf this unlikely to be a great idea. Pavel