Hi Sebastiano, think of it this way: it's not your crystals that decay within 300 degrees, it's you who let them decay that quickly, by not attenuating the beam sufficiently. But now that you know approximately how long they survive (which won't change just because the crystal is bigger, I think), you should figure out how much to attenuate the beam so you can get all the degrees you need. RESIST THE URGE TO BURN FOR RESOLUTION. If your crystal dies, your resolution is useless. If you have enough Se atoms, SAD may be enough to phase, but my rule of thumb is, the fewer Se atoms or the worse the resolution, the more you need to consider getting at least a second wavelength. (Don't stress about the 3rd, not with your crystal dying.) I'd do them sequentially, because if it dies halfway through the second dataset, at least you still have one complete first dataset to do something with maybe. And you want the first to be the one most likely to work as SAD, so I'd do the peak, and then remote (which gets biggest delta-f' from peak). Hope that helps. Above all: DO NOT KILL YOUR CRYSTAL. phx.
_____ From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of Sebastiano Pasqualato Sent: 15 February 2007 14:01 To: [email protected] Subject: [ccp4bb] MAD/SAD data collection strategy Hi all, I'm looking for some advices on some "general hints" on how to carry out a MAD/SAD data collection. We had SeMet crystals that diffracted to ca. 3.5 Angs, with anomalous signal only at ca. 5-5.5 Angs, with diffraction decaying on brilliant beamlines (ID29 or ID23 at the ESRF) in a matter of ca. 300 degrees... Needless to say, that was not sufficient to solve the structure... We do have improved the crystals that look now nicer and bigger, and have some beamtime next weeks both at BM16 and ID29 at the ESRF. Assumed that we do see diffraction higher that 3 Angs, what would people suggest? Collecting first at the high energy remote for a SAD experiment and then going for peak and inflection point, or rather going for the peak first, then remote and ip? I personally would avoid the continuous switch of wavelengths, but I know there are some fans of this technique, either... Let the gurus talk! Any advice is obviously highly appreciated! Thanks in advance, Sebastiano and Claudio -- Sebastiano Pasqualato, PhD IFOM Istituto FIRC di Oncologia Molecolare via Adamello, 16 20139 Milano Italy tel +39 02 574 303 325 fax +39 02 574 303 310
