There are three separate steps in this job:

1. cycle 0: the first BUSTER run uses the starting structural model to phase the structure factor amplitudes; from these phases a 2Fo-Fc electron density map is computed and a prior distribution for the location of the missing atoms is derived; the output of this scaling-phasing only job is LIST.0.html
2. cycles 1-21: the second step is the Maximum Likelihood refinement of the partial structure model, while the distribution for the missing atoms is kept equal to the initial one; data up to 2.0 Å are used for refinement;
3. cycles 22-23: the last step is the Maximum Entropy completion of the missing structure: parameters are varied to modify the density of the initial distribution for the missing atoms; atomic positions and B factors for the atoms in the partial structure are kept fixed to the values they reached at the end of the refinement. Given the degree of imperfection of the partial structure, only data up to 2.5 Å are used for the MaxEnt calculation.

Input Preparation

A few parameters need some attention: see the simple quick input guide

Main items of the output

1. R-factors and Log-likelihood gain plots: the simplest way of checking if the model parameters are indeed improving the fit to the working- and free-set.

   The R values at after the first round of scaling, at cycle 0 are around 40%. By chance the free R is lower than the working R. During refinement (cycles 1-21) the R factors decrease; the final working-set and free-set Rfactors do not differ by less than one percent. The Maximum Entropy completion (cycles 22-23) introduces some overfitting - although still managing to improve the fit to the free set.

   The plot of Log-Likelihood gain vs. cycle number shows an increase of the LLG, starting from the initial value of 0; The working-set LLG is bound to increase because the refinement is driven by maximising the likelihood of the model with respect to the working set; more important is to check that the free-set LLG increases as well. Again, as observed with R factors, some overfitting is evident during Maximum Entropy completion.

2. Correlation coefficients plots: the correlation coefficients curves are a good tool to monitor the improvement in the structural model during refinement and MaxEnt completion; they are independent of the overall scale factor but do depend on the values of the relative scale factors between the partial and missing structures.

   The (F_obs,F_frag) and (F_obs,F_calc) curves only differ at low resolution because in F_calc there are contributions from the solvent and the low-resolution envelope for the missing atoms, while in F_frag only the atoms in the PDB model for the partial structure are taken into account.

   Most importantly, the (F_calc,F_expct) curve depends on the imperfection parameters that parameterise the BUSTER internal error model. The larger the internal estimate for the error on the calculated F, the more this CC curve departs from unity. A comparison between the (F_calc,F_expct) and the (F_obs,F_calc) correlation coefficients curves can inform as to the adequacy of the BUSTER internal error model: if the latter is correct, after the first cycle the two curves should be close to one another.

   The (F_obs,F_obs+d) curve is a measure of the noise on the data vs. resolution. This correlation coefficient is lower than unity when the noise on the data becomes large (typically at high resolution, where the I/s(I) is lowest).

   • It is useful to inspect the Correlation Coefficient curves at cycle 0:

     Cycle 0: After the first round of scaling, the ``observed correlation'' (Fobs,Fcalc) agrees with the ``predicted correlation'' (Fcalc,Fexpct): the initial error model is adequate. The (Fobs,Fcalc) correlation coefficients are better than the (Fobs,Ffrag) at low resolution because of the improvements brought about by solvent and missing atoms models.

     Imperfection B factors: The model imperfection parameter for the partial structure, BImpffrag, decreases while the structure improves. Solvent and missing atoms models are roughly constant during this partial structure refinement (only the missing atoms scale factor and B factor are being refined here), so that the changes in the missing atoms and solvent imperfection parameters are mostly due to the coupling with the partial structure model: as the partial structure contributes less variance ( BImpffrag decreases), the solvent and missing atoms models contribute more ( BImpfsolv and BImpfmiss increase). The Luzzati parameters BLuzzWork, BLuzzFree and BLuzz, are not refined: rather, they are estimated at each cycle from a sigmaA plot, and should also decrease as the structure gets better /more complete.

   • Let's now compare the Correlation Coefficient curves between the end of the refinement (cycle 21) and the end of the MaxEnt completion (cycle 23):

     Cycle 21: At the end of the refinement all the CC values are closer to unity than they were at the beginning: this is due to the improvement of the model for the partial structure. The imperfection Bfactor for the partial structure at the end of refinement has a value around 2 (see curves above): the ``predicted correlation'' (Fcalc,Fexpct) is still in good agreement with the ``observed'' one, (Fobs,Fcalc).

     Cycle 23:After Maximum Entropy completion the (Fobs,Fcalc)curve improves at resolution ranges up to 2.5 Å - due to a better model for the missing atoms. No improvement is brought about to the missing atoms model past the 2.5 Å limit because no Lagrange multipliers were used for those data. The (Fobs,Ffrag) CC curve remains unchanged because no changes are made to the partial structure. The ``predicted correlation'' (Fcalc,Fexpct) is now too pessimistic - during Maximum Entropy completion all imperfection parameters are frozen and not refined anymore.

     Sterochemistry residuals: the plot of the normalised sum of geometry residuals can be used to make decisions as to the value of the X-ray weight. As refinement progresses, and if the X-ray weight is adequate, the normalised sum of observed residuals should tend towards the ideal value of 1 (meaning that ideally every restraint should be obeyed within 1 associated e.s.d.). Final values of the observed residuals higher than unity suggest too loose a geometry - and point to the need for a lower X-ray weight; values lower than unity pertain to too tight a geometry - the X-ray weight could be increased. In this case the refinement hasn't converged yet - so no changes to the X-ray weight are advised. Other useful stereochemistry statistics such as the list of the top violations of stereochemistry restraints are found in the geometry output file;

Final Results

• The coordinate files of the refined partial structure are in the PDB file: shell.01/IF3-C.0.final.pdb.

• The structure factor amplitudes and phases are in the file shell.01/IF3-C.0.final.mtz.

References