Re: [ccp4bb] Unknown positive electron density

[James Holton]

That just looks like bulk solvent to me.  Might be extra dense to the 
the barium, but disordered bulk nonetheless.

If, as you say, the anomalous peak only shows up after you model in a 
strong anomalous scatterrer and not when you don't model in anomalous 
signal, then I suspect all you are seeing is the fabled "model bias".  
Weak signals (like anomalous) are particularly sensitive to this.

It is an interesting test to try:
1) measure the anomalous difference peak height after refinement (I use 
"peakmax" in CCP4)
2) set the Ba atom's f" value to 0, and re-refine to convergence. What 
is the peak value now?
3) set the Ba atom's f" value to twice what you expect theoretically.  
Re-refine. Now what do you see?

Once you've done this, you should have a good feeling for how much of 
your anomalous peak is coming from your model vs from your data.

-James Holton
MAD Scientist

On 8/21/2017 3:05 PM, Betty Chu wrote:
Hi Craig,

The data collection wavelength was 0.92 Angstroms. Since we observe 
anomalous signal for Ba at this wavelength, we would expect greater 
anomalous signal if As were present. There is a possibility for weak 
anomalous signal in this positive density, but the weak anomalous 
signal only shows up if I try to model a Ba in the density. Without 
modelling anything, there is no anomalous signal.

This is what the map looks like after one round of refinement with the 
Ba in the density. But since there are waters that are 1.6 Angstroms, 
1.9 Angstroms, and 2.2 Angstroms away from the Ba, which is smaller 
than the coordination distance between Ba and water, we are skeptical 
of the Ba being there.

https://cbsostorage.chem.umd.edu/owncloud/index.php/s/O4lI2iKvRsQUoRo 
<https://cbsostorage.chem.umd.edu/owncloud/index.php/s/O4lI2iKvRsQUoRo>

Thank you,
Betty

On Mon, Aug 21, 2017 at 5:21 PM, CRAIG A BINGMAN <cabing...@wisc.edu 
<mailto:cabing...@wisc.edu>> wrote:

    What is the data collection wavelength/energy? Would you expect
    significant anomalous diffraction from As at this wavelength?

    On Aug 21, 2017, at 11:37 AM, Betty Chu <chube...@umd.edu
    <mailto:chube...@umd.edu>> wrote:

    Hi Shailesh,

    When I modelled in the Barium ion with octahedrally coordinated
    waters and ran the refinement, the distances from the barium to
    some of the waters ended up being too close (<2.2 Angstroms).
    Also, the positive electron density is connected. If the density
    indicated barium with coordinated waters, would that mean there
    are multiple ones present in the positive density?

    Here are more views of the connected positive density.

    https://cbsostorage.chem.umd.edu/owncloud/index.php/s/j5CgZIJe6NKgUlZ
    <https://cbsostorage.chem.umd.edu/owncloud/index.php/s/j5CgZIJe6NKgUlZ>

    https://cbsostorage.chem.umd.edu/owncloud/index.php/s/YbiNFbT8AW1iHUQ
    <https://cbsostorage.chem.umd.edu/owncloud/index.php/s/YbiNFbT8AW1iHUQ>

    On Mon, Aug 21, 2017 at 12:00 PM, Shailesh Tripathi
    <shailes...@gmail.com <mailto:shailes...@gmail.com>> wrote:

        Looks like Ba2+. Since it exist with coordination number 6 or
        above check what geometry water is following there (trigonal
        bipiramidal or so on). Water might also be shared by symmetry
        related Ba cation.



        Shailesh Kumar Tripathi,
        Phone: 9686289668


        On Mon, Aug 21, 2017 at 9:17 PM, Betty Chu <chube...@umd.edu
        <mailto:chube...@umd.edu>> wrote:

            Yes, I have. The cacodylate ion does not fit well into
            the density.

            On Mon, Aug 21, 2017 at 11:44 AM, Pradeep Pallan
            <pradeeppal...@gmail.com
            <mailto:pradeeppal...@gmail.com>> wrote:

                Did you try modelling in a cacodylate ion (CH3)2AsO2-?

                On Mon, Aug 21, 2017 at 10:19 AM, Betty Chu
                <chube...@umd.edu <mailto:chube...@umd.edu>> wrote:

                    Dear ccp4bb,

                    I am refining a 1.40 Angstrom data set for a DNA
                    oligonucleotide. While the model for the DNA fits
                    very well into the density, there is a patch of
                    positive electron density in the solvent space
                    that we are having trouble with.

                    The screenshot can be viewed through this link:
                    
https://cbsostorage.chem.umd.edu/owncloud/index.php/s/J5cKnOpCC4vb1VC
                    
<https://cbsostorage.chem.umd.edu/owncloud/index.php/s/J5cKnOpCC4vb1VC>

                    In the screenshot, the yellow color is the
                    anomalous map and a barium ion is fitted into
                    density near the positive green electron density.

                    The oligonucleotide was purchased from IDT. The
                    crystallization condition is 15% MPD, 120 mM
                    BaCl2, and 30 mM NaCaC pH 6.4. I have tried
                    modelling Ba2+ with coordinated waters, MPD, and
                    sodium cacodylate into the electron density, but
                    none of those fit well.

                    Any suggestions regarding the identity of this
                    electron density is much appreciated. Thank you!

                    Sincerely,

                    Betty Chu
                    Paukstelis Research Group
                    Department of Chemistry and Biochemistry
                    University of Maryland, College Park




                -- 

                
-------------------------------------------------------------------
                Pradeep Pallan