view article

Figure 2
Mean phase error (blue lines) and weighted mean phase error (grey lines) starting from fragments and the role of model building. Dots show cycles with autotracing, while square points represent no autotracing. (a) Aldose reductase (PDB entry 4lbs; Fanfrlik et al., 2013BB19) at 0.76 Å resolution starting from the phases provided by the Br atom present in the ligand. The deposited structure contains 2567 protein atoms and 687 ligand and water non-H atoms. After 500 cycles of density modification a second run (yellow) iterates autotracing and density modification (ten cycles). (b) Mannose-binding protein A with 167 residues at 1.93 Å resolution (PDB entry 1buu; Ng et al., 1998BB44) from a holmium cation. The darker lines with dots (main-chain tracing) show a faster convergence and better final phases for the same total number of density-modification cycles than when no tracing is applied. (c) PDB entry 1zzk (Backe et al., 2005BB4) at 0.95 Å resolution, where a 82-residue structure can be obtained starting from four helical amino acids; autotracing is not essential but supports convergence. (d) Human myosin 5b (PDB entry 4j5m; Nascimento et al., 2013BB43) at 2.07 Å resolution, where main-chain interpretation is essential to extend from the two starting helices of 17 alanines to an interpretable map; without autotracing the phases deteriorate rather than improve.

Journal logoSTRUCTURAL
BIOLOGY
ISSN: 2059-7983
Follow Acta Cryst. D
Sign up for e-alerts
Follow Acta Cryst. on Twitter
Follow us on facebook
Sign up for RSS feeds