D. Bourgeois1,2, D. Nurizzo3, R. Kahn2, & C. Cambillau3
1 ESRF,
France
2LCCP, IBS, Grenoble, France
3AFMB, CNRS, Marseille, France. email : bourgeoi@esrf.fr
Keywords : macromolecular
crystallography, data processing, profile-fitting, overlap
deconvolution.
We propose a new technique based on profile
fitting for the integration of weak and/or spatially overlapped
2D diffraction patterns [1]. An associated software package with
a GUI interface, named PROW (PRofile fitting for Overlapped and
Weak data) has been developped. Either Laue or monochromatic data
can be processed. Indexation is performed with LAUEGEN [2] or
with the DENZO package [3]. Integration uses box integration with
optimized summation area for strong, non overlapped spots, and
profile fitting with optimized profile fitting area for weak
and/or overlapped spots. Integration areas are dynamically
adjusted to match the spot shape and obtain the best compromise
between minimal bias resulting from statistical fluctuations of
the X-ray background, and maximal use of the information content
in each diffraction spot. Deconvolution of spatially overlapped
spots, a frequent problem in Laue diffraction, and not so rare in
the case of monochromatic data collection, is performed by a
standard least-square procedure. Fourier interpolation of
reference profiles is used to overcome effects of discrete data
sampling. Masks of any shape can be applied to prevent processing
of spurious image areas. Ice spots and obvious outliers are
automatically removed. The program has been tested on several
data sets recorded on image plate and CCD detectors. Results
obtained from very weak Laue data sets, containing as many as 60
% overlapped spots, and from difficult monochromatic data sets
will be shown. Observed improvements in statistical factors
(Rsym, I/SigI, redundancy, completeness) resulted in density maps
and atomic model refinement of superior quality.
[1] Bourgeois, D., Nurizzo, D., Kahn, R. &
Cambillau, C. (1998), J. Appl. Cryst. 31, 22-35.
[2] Campbell, J.W. (1993). Laue Software Suite,
CCLRC Daresbury Laboratory, Warrington, England.
[3] Otwinowsky, Z. (1993). in Proceedings of
the CCP4 Study Weekend, Daresbury Laboratory, Warrington,
England, 56-62.