Hi Rafael, This is progressively getting further off-topic, so we may consider continuing this conversation off-list or under a different subject line.
You are correct that "one should include only what could be seen". However: a) "What could be seen" is not always necessarily a hefty green blob on an Fo-Fc map, and b) Hydrogens are a bit of a special case. As I hinted previously, about 85% of hydrogen atoms (those without rotational degrees of freedom) can be unambiguously assigned to their positions. They are present in actual structures and do contribute to scattering, as including them reduces R — which might align with your definition of "what could be seen" here. Adding hydrogens as a riding model does not introduce additional refinable parameters, while still accounting for their scattering contribution. This is very different from modeling entities such as flexible loops — if you don't see density in the map, there's no reliable way to trace the loop because it may adopt many geometrically plausible conformations. Without data (i.e., map density for the loop), you simply cannot know where to place it — so you may consider not including it in the model. Additionally, including hydrogen atoms makes surrounding atoms aware of them, preventing steric clashes. After all, this is how MolProbity validation works — it automatically adds hydrogens to your model before calculating metrics like Clashscore. By including hydrogens upfront, you proactively reduce the risk of steric clashes, while without them, the only meaningful way to resolve a clash reported by MolProbity would be to add hydrogens and re-refine the model. All the best, Pavel On Mon, Mar 3, 2025 at 3:12 AM Rafael Marques <[email protected]> wrote: > Hi Pavel, > > Interesting! I have always been told that one should include only what > could be seen. In this scenario, I would model hydrogens only if I had a > sub 1 A resolution or if I had Neutron diffraction data. Thanks for the > documentation. > > All the best > > > > ______________________________________________________ > > Rafael Marques da Silva > > PhD Student – Structural Biology > > University of Leicester > > Mestre em Física Biomolecular > Universidade de São Paulo > > Bacharel em Ciências Biológicas > Universidade Federal de São Carlos > > phone: +44 07861 273773 > > * "A sorte acompanha uma mente bem treinada"* > *________________________________________________* > > ------------------------------ > *De:* Pavel Afonine <[email protected]> > *Enviado:* domingo, 2 de março de 2025 22:10 > *Para:* Rafael Marques <[email protected]> > *Cc:* [email protected] <[email protected]> > *Assunto:* Re: [ccp4bb] unknown density > > Hi Rafael, > Regarding your question about hydrogens: it's a good practice to include > them using the riding model towards the end of refinement, regardless of > resolution. Hydrogens make up about half of the model atoms, contribute to > the scattering, and about 85% of them have geometrically unambiguous > positions based on their parent atoms. Including hydrogens also helps > achieve more sensible model geometry by improving non-covalent interactions > and reducing steric clashes. > > Some notes related to this: > > https://phenix-online.org/phenixwebsite_static/mainsite/files/newsletter/CCN_2012_01.pdf#page=18 > > All the best, > Pavel > > On Sat, Mar 1, 2025 at 8:20 AM Rafael Marques <[email protected]> > wrote: > > Hi Michael, > > It looks like PEG to me too. Have you used a smaller PEG as > cryoprotectant. Such as PEG 200 or PEG400? Also, as an off topic thing, why > are you refining hydrogens at this resolution? > > Best wishes > > > > ______________________________________________________ > > Rafael Marques da Silva > > PhD Student – Structural Biology > > University of Leicester > > Mestre em Física Biomolecular > Universidade de São Paulo > > Bacharel em Ciências Biológicas > Universidade Federal de São Carlos > > phone: +44 07861 273773 > > * "A sorte acompanha uma mente bem treinada"* > *________________________________________________* > > ------------------------------ > *De:* CCP4 bulletin board <[email protected]> em nome de Kennedy, > Michael <[email protected]> > *Enviado:* quinta-feira, 27 de fevereiro de 2025 19:57 > *Para:* [email protected] <[email protected]> > *Assunto:* [ccp4bb] unknown density > > Hello All, > > We recently solved the structure of a protein at 1.8A. There is a large > unknown density (see attached) at the interface between two protein > molecules. > > The crystallization buffer was 0.2 M Li Sulfate, 0.1 M Tris,25% PEG 4000, > 10% glycerol. > The protein purification buffer was 250 mM NaCl, 20 mM Tris, 10% glycerol, > 300 mM imidazole. > > Does anyone recognize this ligand or have an idea about what it might be? > > thanks > > Michael A. Kennedy, PhD > > Eminent Scholar and Professor > Department of Chemistry and Biochemistry > 106 Hughes Laboratory > Miami University > 651 East High Street > Oxford, OH 45056 > > phone: 513-529-8267 > fax: 513-529-5715 > email: [email protected] > webpage: http://chemistry.muohio.edu/kennedy/ > google scholars citations: > https://scholar.google.com/citations?user=7rWzfjkAAAAJ > NCBI My Bibliography > https://www.ncbi.nlm.nih.gov/myncbi/1L3_kwfzxCU/bibliography/public/ > Lab Twitter : @Kennedy_Lab > > ------------------------------ > > 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 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/
