For predicting the scattering curves, AXES accepts any PDB-formatted file with the atomic coordinates, but currently is restricted to the following atoms: C, O, N, H, S, P, Fe, Na, K, Cl, Cu, Zn, Mn, Mg, Ca, Pb. The PDB file can include multiple structures as long as they are separated by lines with the END or ENDMDL keywords. In that case, the fit will be performed with respect to the ensemble-averaged predicted scattering functions but the curves predicted from the individual models will be reported separately. Results calculated for an atomic model will be slightly more accurate if 1H coordinates are included in the model. For proteins, 1H coordinates can be added by the PDB Utility Server or by any other program such as REDUCE or MOLMOL.

AXES can calculate best fits between experimental data and individual structures or ensembles of structures, while minimizing the use of adjustable parameters in this process. The latter include a q-independent offset due to the detector dark current fluctuations or incoherent background scattering and an adjustable scaling factor for buffer subtraction, which mimic the experimental uncertainties most commonly encountered in experimental data, as well as the increased electron density of the solvation layer. The limited use of adjustable parameters makes the fitting procedure particularly robust in distinguishing correct structures from decoys.

Reference: A. Grishaev, L.A. Guo, T. Irving, and A. Bax: Improved fitting of solution X-ray scattering data to macromolecular structures and structural ensembles by explicit water modeling. J. Am. Chem. Soc. 132, 15484-15486 (2010) 381.pdf 381_supportinfo.pdf