Hi Jacob,
As one of the (sadly) remaining co-founders of the MASC method, I can confirm Pierre's comments. MASC signals are VERY strong at very low resolution, and then it peters out with increasing resolution until ca 5 Angstroms. When we collected MASC data at LURE and the ESRF in the 1990s, we observed the very strong signals at low resolution and in some cases ordered sites. In the scaling and merging steps of data processing, we often observed the weaker Friedel & Bijvoet mates discarded by scaling programs. This motivated Richard Kahn to contact Wolfgang Kabsch to ask for the option that Friedel & Bijvoet pairs be considered as separate reflections in scaling. Since then, I know that XDS handles this nicely. Other scaling programs also handle the Friedel & Bijvoet mates separately, but I do not know their details.
Could MASC signals perturb the anomalous signal in SAD or MAD experiments? If the lowest resolution reflections are required for phasing, then I would say POSSIBLY YES. Consider that direct methods uses the lowest resolution reflections in the phase relationships, and that often poorly measured low resolution reflections will compromise the solving for the sub-structure. (For example, prior to the arrival of the higher dynamic range of the PAD detectors, the lowest resolution reflections could easily saturate on a CCD-based detector.) Thus reflections exhibiting strong MASC signals could perturb the HA structure solution process if insufficient phase relationships are generated.
MASC signals become significant when the anomalous scattering atoms are present at molar levels for L-edges and multi-molar levels for the K-edge. A crystallographer might wish to consider MASC effects if their crystals contain 4M sulphate for a sulphur SAD experiment, or 4M selenate for selenomethionine-substituted protein crystal for SAD or MAD data collected at or above the Se K-edge.
Finally, yes, the MASC method could be observed at long wavelengths on beamlines such as I23 at DLS. The only remaining unresolved issue is the phase problem, because the MASC method produces amplitudes of the macromolecular envelope, which does not fulfil the criteria of "atomicity" required in all (?) crystallographic structure solution methods. This was one of Richard Kahn's research topics before his tragic death a few years ago.
I hope this answers some of your questions. It has revived many questions (and good memories) for me.
Cheers,
Bill
________________________________________
From: CCP4 bulletin board [[log in to unmask]] on behalf of Keller, Jacob [[log in to unmask]]
Sent: 24 May 2016 14:57
To: [log in to unmask]
Subject: Re: [ccp4bb] Anomalous Scattering of Bulk Solvent
For example: with NH4SO4 at ~2.5-3 M, that's about one sulfate per 20 waters, and the sheer amount of anomalously-scattering material would have a significant effect on phases, like a continuum of thousands of poorly-localized S-sites with 0.05 occupancy. Roger Fourme's work indeed seems to prove this--he used stronger scatterers and multiple wavelengths, but the nature of the effect would be the same.
I wonder whether this effect could be used in conjunction with the new low-energy native SAD beamline at Diamond? Low-resolution spots should be readily measured with such long x-rays.
JPK
-----Original Message-----
From: LEGRAND Pierre [mailto:[log in to unmask]]
Sent: Tuesday, May 24, 2016 5:32 AM
To: Keller, Jacob; [log in to unmask]
Subject: RE:Anomalous Scattering of Bulk Solvent
Dear Jacob,
As you mentioned in your message, because this is a solvent contrast method (I suppose you are referring to the MASC method from Roger Fourme et al), this effect can only be exploited at quite LOW resolution.
At higher resolution, if you have ordered sites, this effect will most probably be negligible.
Also, given the low anomalous form factor value of the elements you are mentioning and their probable strong dilution in the solvent, it guess that most S-SAD cases will be more adequately characterized by having extremely-weak-anomalously-scattering solvents.
Now, if you still fear that some software might be confused, you can always try to apply a low-resolution cut-off.
Best regards
Pierre
PROXIMA-1 @ SOLEIL
________________________________________
De : CCP4 bulletin board [[log in to unmask]] de la part de Keller, Jacob [[log in to unmask]] Envoyé : lundi 23 mai 2016 20:19 À : [log in to unmask] Objet : [ccp4bb] Anomalous Scattering of Bulk Solvent
Dear Crystallographers,
I have seen that some have exploited highly-anomalously-scattering solvents in protein crystals to define envelopes at low resolution, which is an interesting technique, but I was wondering whether a similar but deleterious effect might occur in crystals containing the common molar concentrations of sulfate, chloride, phosphate or others. More precisely, this might represent a problem in phasing from weak, ordered scatterers in crystals so grown, since the low-resolution bulk solvent would be giving different information from the ordered sites, and this might confuse the software. Is account taken of this possible effect in the HA/phasing software? Would it be worth it?
I thought this might be significant since there is a growing interest in using S-SAD, for example.
JPK
*******************************************
Jacob Pearson Keller, PhD
Looger Lab/HHMI Janelia Research Campus
19700 Helix Dr, Ashburn, VA 20147
email: [log in to unmask]
*******************************************
|