Hi Eleanor, My understanding about negative holes in difference maps is that it is down to incorrectly modelled bulk solvent. I have seen this in structures where there are voids inside a protein that are perhaps in reality empty, but get modelled as bulk solvent, and so you get a negative difference peak filling the void.
I am more than happy to be corrected, though. Best, Dave -- Dr David C. Briggs CSci MRSB (he/him) Principal Laboratory Research Scientist Signalling and Structural Biology Lab The Francis Crick Institute London, UK Working hours: Mon-Fri 0900-1700 == about.me/david_briggs<https://about.me/david_briggs> | OrcID<https://orcid.org/0000-0002-9793-7339> | Google Scholar <https://scholar.google.co.uk/citations?user=DRKG5KwAAAAJ> ________________________________ From: CCP4 bulletin board <[email protected]> on behalf of Eleanor Dodson <[email protected]> Sent: 25 July 2024 07:10 To: [email protected] <[email protected]> Subject: Re: [ccp4bb] Presence of negative density in Sim Omit map External Sender: Use caution. Hmm - I cant quite understand your map - that density is not for the lysine ? It looks like a well ordered PHE contoured at quite a high level? As Edward suggests - if you omit a well ordered feature the resultant difference maps often show high positive density and a surrounding complementary "sag" of negative density - not really something to worry about.. But I wonder why you don't just do a bog standard difference map? Set the occupancies of the loop you want to omit to 0.00 - do a few cycles of refinement and see what comes back? A prediction - the well ordered features will show up loud and clear and your surface LYS wont! High resolution structures show they often are in multiple conformations, which would be hard to model at 2.1A. And just a thought about negative features in difference maps. I often see random "holes" which don't seem to have any logical explanation.. I sloppily put them down to "data defects - missing data? poorly measured data? etc - but does anyone have a more satisfactory explanation? Or dont other people see them?? Eleanor On Thu, 25 Jul 2024 at 01:48, Edward A. Berry <[email protected]<mailto:[email protected]>> wrote: Does your difference map have mean value ~zero (over 1 ASU or cell)? If maps are constructed by Fourier transform without the 0 0 0 reflection, they have mean of zero (because the mean of a sinusoid over one period is zero). That means that any time you add positive (difference) density, which raises the mean value of the absolute map, the Fourier map has to sink down a little to bring that new mean value to zero. Now if your data were complete except for the 0 0 0 reflection, this would move the floor down everywhere by a small amount to balance the large increase at the local peak, and it would not go below the negative contour limit. But without the ultra-low resolution amplitudes the FT cannot make a constant offset, and instead you get sagging around the peak, like a water-bed sagging around the spot where a child (the peak) is standing. This could lead to the sagging part going below the negative contour and giving you the red density. However I would not have expected the negative density to be quite so localized around the peak - depending on your low-res cutoff. As John Bacik suggested you should check for a mis-modeled part of a symm-related molecule, but if that were the case, you should see the red density even before omitting the loop, and you should see the red density somewhere else on your model. If it just appeared after omitting the loop, it could be due to the map sagging under the weight of the peak, and should go away as soon as the residue is replaced. eab Renuka Kadirvelraj wrote: > Hi CCP4bb, > We would greatly appreciate your advice regarding an odd problem that has > cropped up with one of our > crystal structures. We have a protein structure in space group P6(3)22 at 2.1 > A resolution and in > the final refinement stages with Rwork of 0.20 and Rfree of 0.22. When we ran > simulated annealing to > rebuild an outlier Lysine residue in a loop (using Phenix), the Sim Omit map > showed the positive > density expected from the omission of the loop (colored green in attached > pics). However, flanking > the positive density, there is a complimentary negative density (in red). We > are puzzled as to the > cause of the appearance of the negative density. Can you help us? > Many thanks, > Renu > > --------------------------------------------- > Renu Kadirvelraj, Ph.D. > Research Scientist > A428, Biochemistry and Molecular Biology > 120, East Green Street > University of Georgia > Athens, GA 30602 > Tel: (706) 583 0303 > > ---------------------------------------------------------------------------------------------------- > > 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<http://www.jiscmail.ac.uk/CCP4BB>, a mailing list hosted by www.jiscmail.ac.uk<http://www.jiscmail.ac.uk/>, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/ ________________________________ To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 The Francis Crick Institute Limited is a registered charity in England and Wales no. 1140062 and a company registered in England and Wales no. 06885462, with its registered office at 1 Midland Road London NW1 1AT ######################################################################## 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/
