PDB COMMUNITY FOCUS: KIM HENRICK, E-MSD
Kim Henrick is currently the group leader of the European
Bioinformatics Institute (EBI) Macromolecular Structure Database
(E-MSD). He was born in Australia and received his PhD at the
University of Western Australia in Crawley, Western Australia. He was
a post-doctoral fellow at the Imperial College of Science and
Technology in London, The Cambridge Centre for Protein Engineering in
Cambridge, and the National Institute for Medical Research in
London. Kim also spent some time working on the CCP4 project and as the
senior scientific officer at the UK Science and Engineering Research
Council (SERC) Laboratory in Daresbury. His interest in things
structural has spanned the gamut from solving the structure of small
organic molecules (reacting thionyl chloride with tricyclo
[3,2,1,02,4]oct-6-en-ols to developing web-based systems for the
deposition and validation of high-resolution electron microscopy
macromolecular structure information. He has been with the European
Bioinformatics Institute (EBI) since 1996, and became the head of the
Macromolecular Structure Database (E-MSD) in 2001. Kim is also
actively involved with the newly formed World Wide PDB (wwPDB).
Q: You started out doing small molecule crystallography and are now
involved in the protein world. How and why did this transition take
place for you?
A: The institute where I was working in the 1980's was in financial
difficulties so I took voluntary redundancy and moved into protein
crystallography. This was initially in bewilderment coming from
2-theta as a measure of resolution to Angstroms and to the
realization that a B-factor in the protein world had no physical
Q: As macromolecular crystallography moves into the age of structural
genomics do you see parallels with changes that have taken place in
small molecule crystallography over the years?
A: In 'small molecule crystallography' (what a dreadful expression) we
had George Sheldrick. The release of SHELX76 changed the world in
crystallography from struggling with programs with card image formats
(SHELX76 was testing in 1975 -- the same year Microsoft was founded
by Bill Gates and Paul Allen). This was followed closely by Digital
Equipment Corporation introducing the VAX 11/780 (1978). These two
events made crystallography relatively easy. With George's SHELXTL, he
put into a single package data reduction, phasing, refinement,
graphics, and report generation. Protein crystallographers at present
have a status that has been lost by most chemical crystallographers
although both are collecting data with the same machines (cryo-cooling
and image plates). Currently data collection and structure solution in
chemistry is usually carried out by departmental service
crystallographers. The speed of data collection and structure solution
for protein ligand complexes or mutant analysis is now much the same
as for chemical crystallography. If the various high throughput
projects being undertaken, such as BIOXHIT in Europe, achieve their
aims it should soon be possible for new protein structures to be
solved at similar rates, (and now 'we' have George Sheldrick as well,
especially with SHELXD and SHELXE).
Q: What is the scope of the MSD project?
A: The aim of the European Macromolecular Structure Database, MSD, is
to support deposition of new structures determined by X-ray, NMR and
cryo-3D-electron microscopy techniques, to provide storage and
organization of the structural data, and to support search and
analysis tools to query the data, i.e. exactly the same as the RCSB
Q: What is the nature of your interactions with the RSCB PDB and PDBj?
A: The members of the wwPDB have identical aims in managing,
processing and providing publicly accessible structural data. Not just
the best we can do but really in the right way. The groups consist of
people dedicated to this end of providing a global service for 3D
structure data and we each know our individual processes are the best
approach - which leads to resolvable niggling tensions at
times. However these separate developments give an invaluable means of
cross checking PDB entries, which would otherwise be missed. The
members work under pressure from scientists who want the data in a
form they require now, while in part being unsympathetic to the
standardization of their structures.
Q: What do you think wwPDB should accomplish?
A: The wwPDB should reach the situation where the PDB is recognized as
a global resource and its role defined by the partners and their
scientific advisory boards. This will take considerable effort from
the partners who are each restrained by the specific demands of their
funding agencies. However, the function of the PDB should be seen as
not the sole prerogative of the USA funding agencies.
Q: Where do you see structural biology evolving over the next decade
and how will the PDB need to change to keep pace?
A: Genome projects and sequence-oriented bioinformaticians see their
work as dealing with molecular biology. However, on its own, the DNA
sequence tells us little about what the genome does or how it works.
DNA sequences code for proteins but proteins are only functional once
they have folded up into a unique 3D structure. Similarities and
differences found between gene sequences can only really be understood
once the 3D structure of a member of a homologous family is known. The
PDB currently handles 3D structure data for proteins from the
experimental techniques of X-ray crystallography and NMR
spectroscopy. Significant 3D structural information is given from
cryo-3D-electron microscopy and tomographic reconstructions. The PDB
perhaps should expand to include all 3D experimental data and
concentrate on data integration with other biological databases.