Reliable structural interpretation of small-angle scattering data from bio-molecules in solution - the importance of quality control and a standard reporting framework

The theory underpinning small-angle scattering (SAS) and its application to the study of polymer structures in solution have been known for over 80？years [1]. Early in the development of SAS methodology it was recognised that biomolecules were ideal candidates for structural characterisation because they can be prepared as a solution of identical, mono-disperse particles as a result of the fact that a given polypeptide or polynucleotide of defined sequence folds to form a well-defined structure [2]. However, until the last two decades only a few, readily prepared systems were studied by SAS. The modern ‘explosion’ in the use of the technique for structural biology was made possible by the advent of molecular biology and modern biochemistry tools for milligram-scale purification of proteins and polynucleotides for sample preparation; improvements in performance and availability of SAS instrumentation; and modern computing with easy-to-use software for data processing and interpretation. Perhaps the most attractive development to the structural biologist was the ability to generate coordinate files representing 3D structures, whether they be ab initio molecular envelope (bead) reconstructions, or atomic models based on the refinement of rigid body domains obtained by another method (e.g. crystallography, NMR, or high confidence homology models based on experimental structures) [3]. While such algorithms can generate visually appealing graphics, the inherently low information content of a 1D small-angle scattering profile from unoriented macromolecules in solution means that without careful evaluation of sample and data quality, as well as a good understanding the information limits of the data, the structures obtained can be misleading or simply wrong.For the results of biomolecular-SAS studies to be relied upon by the community, a standard reporting framework is needed specifying what details of the experiment and its interpretation must be provided. A similar step w