[phenixbb] David Sayre

[Paul Adams]

On behalf of the Phenix development team:

We regret deeply the passing of David Sayre on Feb. 23, 2012. David Sayre made profound contributions to crystallography and scientific computing that continue to have a major impact decades later.

An obituary summarizing his many achievements was posted by Physics Today (http://www.physicstoday.org/1.2732466). In crystallography, Sayre’s fundamental insight into the implications of the atomic nature of crystal structures and the relationships between crystallographic structure factors allowed him to develop in 1952 the brilliantly simple Sayre equation relating each structure factor to all other structure factors.  The Sayre equation has had a well-known yet greatly under-appreciated significance in small-molecule crystallography, both in its original form and in the equivalent probabilistic forms implemented later as part of direct methods. These methods have profoundly affected macromolecular crystallography in the last decade because of their astounding power for determination of heavy-atom substructures. The use of direct methods has made possible solution of the positions of as many as 160 selenium atoms in large macromolecular structures.  To appreciate the feat that this represents, it is useful to consider that many macromolecular crystallographers were initially hesitant to use the seleno-methionine approach at all for structures that had more than about 10 selenium atoms for fear that the substructure could not be solved.

Sayre’s recent work on phasing X-ray diffraction data from single large objects such as a yeast cell and single small objects such as a single protein molecule was groundbreaking in a different way. This work is revolutionary because of its boldness, and because of the elegantly simple approach to phase determination that Sayre developed.  Once again Sayre took a fundamental real-space feature of the diffraction image and used it to obtain phase relationships. In this case the diffraction forms a continuous pattern from a single object, and the feature is the lack of contrast outside of the object. While the idea of using a mask of the object (the “support”) to define phases has been used for some time in many contexts, it was far from obvious that this could be done with a single cell or with a single protein molecule.  Sayre and his colleagues systematically showed theoretically and then experimentally that each step in this process can be carried out and that the required signal can be extracted to obtain an X-ray image of microdots and later even a yeast cell.  Others have followed up his ideas for phasing data from a single protein molecule. This work was of fundamental and transforming importance because few would have believed that the ideas were remotely feasible, and now they have a sound theoretical basis and experimental demonstrations, allowing anyone to think of new and powerful ways to use these approaches.

The breadth of Sayre's intellectual spectrum reached far beyond crystallography. He was also a member of the IBM team that developed the original Fortran compiler, the first high-level programming tool (see http://www.ibm.com/ibm100/us/en/icons/fortran/team/ for more details). Currently, 55 years after the first commercial release of Fortran, this language is still in wide use. The history of crystallographic computing, and scientific computing in general, cannot be separated from the success of Fortran, made possible by the amazing IBM team that included Sayre.

In spite of his great accomplishments, David Sayre was modest and approachable in person, and he was generous in sharing credit with his colleagues. He will be greatly missed.

-- 
Paul Adams
Deputy Division Director, Physical Biosciences Division, Lawrence Berkeley Lab
Division Deputy for Biosciences, Advanced Light Source, Lawrence Berkeley Lab
Adjunct Professor, Department of Bioengineering, U.C. Berkeley
Vice President for Technology, the Joint BioEnergy Institute
Laboratory Research Manager, ENIGMA Science Focus Area

Building 64, Room 248
Tel: 1-510-486-4225, Fax: 1-510-486-5909
http://cci.lbl.gov/paul

Lawrence Berkeley Laboratory
1 Cyclotron Road
BLDG 64R0121
Berkeley, CA 94720, USA.

Executive Assistant: Louise Benvenue [ LBenvenue at lbl.gov ][ 1-510-495-2506 ]
--