[sharp-discuss] Rcullis(anomalous) for SAD data?

[Edward MITCHELL]

    Dear Sharpers,

    Question: what is the definition of Rcullis(anomalous) for single 
wavelength
    anomalous data used on it's own with no native data (i.e. "SAD"/"SAS")?

    The mmcif defintions at Rutgers give:

                           sum| |Fph+~obs~Fph-~obs~ - Fh+~calc~ - Fh-~calc~| |
            R~cullis,ano~ = ---------------------------------------------------
                                   sum|Fph+~obs~ - Fph-~obs~|

               Fph+~obs~  = the observed positive Friedel structure factor
                            amplitude for the derivative
               Fph-~obs~  = the observed negative Friedel structure factor
                            amplitude for the derivative

               Fh+~calc~  = the calculated positive Friedel structure factor
                            amplitude from the heavy atom model
               Fh-~calc~  = the calculated negative Friedel structure factor
                            amplitude from the heavy atom model

    If I try and write this more cleanly:
    Rcullis(anom) = E  |  Fph+(obs)Fph-(obs)  -   Fh+(calc)  - 
 Fh-(calc)   |   /   E  | Fph+(obs) -   Fph-(obs)  |

    But by analogy with the normal Rcullis:
    Rcullis = E | | Fph(obs)   +   Fp(obs)  |  -  | Fh(calc)| |  /  E | 
Fph(obs) - Fp(obs) |

    Rcullis(anom) surely should be:
    Rcullis(anom) = E  | | Fph+(obs)   +   Fph-(obs) |  -  | Fh+(calc) 
 -  Fh-(calc) | |   /   E  | Fph+(obs) -   Fph-(obs)  |  

    Which (if any) is correct (i.e. what SHARP uses in it's calculations)?

    thanks in advance!

    Ed.

    (the same question then applies to Phasing Power of course!)