Hello,
This review also contains info relevant to this discussion: https://doi.org/10.1016/j.sbi.2019.04.006 Figure 1 shows more examples of electron versus x-ray atomic scattering factors for various elements and their charged states. And this paragraphs is very relevant to Matthew's questions: "X-ray crystallographic electron density maps accurately represent the positions of atomic centers because peaks in most maps are centered on the nuclei. This approximation breaks down at very high resolution (<1 Å) where electron orbital structure becomes resolved and electron density peaks are distributed around nuclei [11]. In contrast, peaks in Coulomb potential maps [= cryoEM maps] are not necessarily centered on atomic nuclei, particularly when atoms carry a net charge. Unlike X-ray scattering, electron scattering has a strong charge dependence (e.g. compare O and O− in Figure 1a). This phenomenon can result in deviations of Coulomb potential map features from atomic centers that can confound accurate model building using current methods." So, it seems like difficult agreement between a cryoEM map and ideal stereochemistry is to be expected. The review goes on to cite a series of papers discussing this point in more details: https://doi.org/10.1002/pro.3060 https://doi.org/10.1002/pro.3169 https://doi.org/10.1002/pro.3198 : this paper explains how to convert a Coulomb potential map (= an unfiltered/unsharpened cryoEM map) into a charge density map, which is presumably easier to interpret and in better agreement with ideal geometry? I have not yet tried this, and in fact have not yet read this series of papers closely. But this seems very interesting, I will try this with the atomic resolution apoF cryoEM maps from 2020 if/when I find the time. https://doi.org/10.1002/pro.3093 I hope this helps, Guillaume ________________________________ From: CCP4 bulletin board <[email protected]> on behalf of zbyszek <[email protected]> Sent: Tuesday, July 29, 2025 1:56:18 AM To: [email protected] Subject: Re: [ccp4bb] Restraining metal sites in medium resolution cryo-EM I believe much more work is needed to understand charge effects in cryoEM. An interesting paper (https://www.nature.com/articles/s42004-023-00900-x) shows in Fig. 3 that these effects for oxygen extend to medium resolution. However, a complete understanding also requires consideration of radiation effects, as radiation exposure rapidly redistributes charges in addition to causing damage and decay. The charge on a given atom may already be affected by exposures in the kGy range, meaning that even the first frame of a cryoEM movie is impacted. The first frame typically corresponds to a dose of ~3–4 MGy (~1 e⁻/Ų). Total doses in cryoEM are high (on the order of hundreds of MGy) so radiation-induced damage to sensitive groups, including those that coordinate metal cations, is significant. Because a cryoEM structure, even after up-weighting the early frames, represents a partially damaged state, the coordination sphere, if left unrestrained, is difficult to interpret. Zbyszek On 2025-07-28 18:31, James Holton wrote: Yes, but formal charge effects are not implemented. Last time I checked? -James Holton MAD scientist On 7/25/2025 11:01 AM, Dominika Borek wrote: I don't think there's anything particularly difficult about scattering factors for cryoEM, but the patterns of dependences on atomic number differs from that in X-ray crystallography. For instance, in X-ray maps, oxygen atoms, common interactors of metal cations via carboxylic acids and/or water molecules, appear "stronger" due to this difference. Thus, at the same nominal resolution of ~2–3 A, the coordination sphere in cryoEM maps looks less well-defined than in X-ray crystallography maps (even though information content is the same). I find the default restraints insufficient, so I apply restraints manually in both Refmac/Servalcat and Phenix, e.g., I define Mg2+ octahedral geometry with proper distances in a mini script and include it in the refinement run. In other words, I vote for tighter geometry. D. Dominika Borek, Ph.D. *** UT Southwestern Medical Center 5323 Harry Hines Blvd. *** Dallas, TX 75390-8816 214-645-9577 (phone) *** 214-645-6353 (fax) From: CCP4 bulletin board <[email protected]><mailto:[email protected]> On Behalf Of Frank von Delft Sent: Friday, July 25, 2025 10:10 AM To: [email protected]<mailto:[email protected]> Subject: Re: [ccp4bb] Restraining metal sites in medium resolution cryo-EM Isn't there something about scattering factors being ... difficult... for CryoEM reconstructions? Which I imagine would cause issues in refinement. Frank On 25/07/2025 12:35, Matthew Snee wrote: Hi Everyone Id let to get the communities opinions on metal sites in Cryo-EM models. In X-ray structures the geometry of metal coordination (when restrained correctly) looks pretty good, even in low resolution structures, and at very high resolution you may wish to avoid applying restraints so that the precise coordination distances/angles can tell you something about the physical chemistry of the system (I.E reduction state of catalytic metals in chemical biology). I have found that EM structures between 2-3Å which are good enough to see individual features (un ambiguous sidechain rotamers and coordinated waters in the metal complex) refinement outputs don't look nearly as nice. My question is this,: Is this just "the way it is" because the optical resolution and true atomic resolvability is different between X-ray and EM. The constraints of the crystal might help to collapse the conformational landscape somewhat, and solvent flattening might repress (unidentifable) minor states that would distort the real-space fit, thus improving the convergence between ideal geometry and fit to the density. EM structures that have very high resolution FSC cutoffs can still be distorted by minor anisotropic flexibility in one or more particular direction, there are ways to isotropise maps, but I dont really like refining against these. The other option is that the restraints need to be up-weighted so that they have more parity with the protein bond restraints and other targets that one might use. Basically, are these structures where the coordination geometry is almost certainly not completely correct, likely to be better, more useful, or closer to the "true structure" or is it better to enforce the geometry more strictly than you would for an equivalent X-ray structure in order to achieve the most likely sensible cluster. I dont think either approach is "wrong" because you could make an argument for building what you see, only building the links that are directly suggested and restraining them as best you can, I'm not a metal cluster expert by any means, but I do know that the coordination number, and bond length/angles are quite well known. I know this is potentially more of a CCPEM/Phenix query, but X-ray people are generally more involved in the physical chemistry side (I know not always 🙂!), Best wishes Matthew Snee, PhD Post-doctoral Research Associate The Baldock Lab | Michael Smith Building C3.214 | Wellcome Trust Centre for Cell-Matrix Research |Division of Cell Matrix Biology & Regenerative Medicine| School of Biological Sciences| Faculty of Biology, Medicine and Health, University of Manchester, Oxford Road, Manchester M13 9PT Lab Tel: +44 (0)161 306 2869 [cid:[email protected]][cell-matrix-research] ________________________________ To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 ________________________________ To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 ________________________________ To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 ________________________________ To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 ________________________________ To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 VARNING: Klicka inte på länkar och öppna inte bilagor om du inte känner igen avsändaren och vet att innehållet är säkert. CAUTION: Do not click on links or open attachments unless you recognise the sender and know the content is safe. När du har kontakt med oss på Uppsala universitet med e-post så innebär det att vi behandlar dina personuppgifter. För att läsa mer om hur vi gör det kan du läsa här: http://www.uu.se/om-uu/dataskydd-personuppgifter/ E-mailing Uppsala University means that we will process your personal data. For more information on how this is performed, please read here: http://www.uu.se/en/about-uu/data-protection-policy ######################################################################## To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
