Ethan and Phil, Your responses about i) normalization based on minimizing the average differences between Fobs and Fcalc and ii) the smearing of electron density by various physical phenomena, makes sense. The smearing cannot easily be accounted for and having density, where it should not be, can be interpreted a couple different ways.
Thanks for the quick responses. Cheers, Hunter On Tue, Feb 21, 2017 at 5:46 PM, Ethan A Merritt <[email protected]> wrote: > On Tuesday, 21 February, 2017 16:53:06 Hunter Moseley wrote: > > Is there a straight-forward way to estimate the amount of missing > electron > > density that a particular protein structure is missing based on the > > difference between Fo and Fc? > > Short answer: no. > > > It appears that the normalization of the Fc due to the employing of a > > maximum entropy method that keeps Fo and Fc comparable to the standard > > deviation of Fo would make this difficult. > > Maximum entropy is not the issue. Your Fobs have no intrinsic scale. > They are not a count of photons-per-second or a fractional intensity of > the direct beam. In order to refine a model you must introduce an > ad hoc scale factor to make Fobs approximately equal to Fcalc on average. > "On average" is a somewhat sloppy description (that's where use of > maximum likelihood weighting may come in) but in the end we normally > calculate and display difference-density maps such that the mean > difference density is zero. You won't end up with a map that is > overall negative even if your model has spurious extra bits and you > won't end up with a map that is overall positive even if your model > is missing large chunks. The best you can hope for is that your > missing chunks will manifest in the map as connected blobs of positive > difference density balanced out by diffuse/disconnected regions of > negative difference density elsewhere. > > Longer answer: still no. > > The distribution of differences between Fobs and Fcalc can provide an > estimate of how good/bad wrong/right your current model is. > But that still doesn't tell you whether the current model is bad because > pieces are missing or bad because existing pieces are in the wrong place. > > > Ethan > > > > Or am I missing something? > > > > Cheers, > > Hunter > > > > > > -- > Ethan A Merritt > Biomolecular Structure Center, K-428 Health Sciences Bldg > MS 357742, University of Washington, Seattle 98195-7742 > > -- Hunter Moseley, Ph.D. -- Univ. of Kentucky Associate Professor, Dept. of Molec. & Cell. Biochemistry / Markey Cancer Center / Resource Center for Stable Isotope Resolved Metabolomics Not just a scientist, but a fencer as well. My foil is sharp, but my mind sharper still. --------------------------------------------------------------- Email: [email protected] (work) [email protected] (personal) Phone: 859-218-2964 (office) 859-218-2965 (lab) 859-257-7715 (fax) Web: http://bioinformatics.cesb.uky.edu/ Address: CC434 Roach Building, 800 Rose Street, Lexington, KY 40536-0093
