|
|
1993 Media Coverage
- 1993: An article in the IUCr Newsletter, vol. 1, no. 3,
states that three papers by
Dr. Russ Miller,
Charles M. Weeks, David Langs, and Herb Hauptman reported
on improvements in the "Shake-and-Bake" method for determining structures
using the minimal function.
- December, 1993: An article in Interface entitled
"Research Team Succeeds with the Aid of Supercomputers" states that
a team of local researchers developed a procedure called "Shake-and-Bake"
and have implemented it on parallel supercomputers to achieve spectacular
successes. The article states that
Dr. Russ Miller
and Dr. Charles M. Weeks of the Medical Foundation of Buffalo have
succeeded in solving difficult molecular structures in a matter of hours.
The article goes on to describe the basics of x-ray crystallography and
direct methods. It then discusses the Shake-and-Bake solution, developed
by Miller, Weeks, Nobel Laureate Herbert A. Hauptman, and other members
of the Medical Foundation. The article states that the procedure is
computationally intensive and that they have taken advantage of resources
at Thinking Machines Corporation and at the Pittsburgh Supercomputing Center.
the story concludes by stating that Dr. Miller is currently working on a
user-friendly version of the program, which he hopes to release by the
end of the year.
- September, 1993: An article in Impact entitled
"MFB Scientists are a Hit at National Meeting" states that
two posters by Drs. Weeks, Miller, and Hauptman were displayed that
described the first application of their computer analysis technique, the
minimal principle, to bona fide proteins. In fact, the article states that
the remarkable success of SnB caused
Dr. George Sheldrick, the winner of the ACA's Patterson Prize for 1993,
to abandon his intended address in order to draw the attention of the
600 people attending his lecture to the importance of this new research.
Dr. Sheldrick expressed his conviction that the minimal principal constitutes
a critical breakthrough in methods of x-ray crystal structure determination.
The action of Sheldrick was without precedent in the 45 year history of
the ACA.
- 1993: An
article (only a small excerpt is available on-line)
in Projects in Scientific Computing,
from the Pittsburgh Supercomputing Center, which also includes the
photo of Charles M. Weeks, Russ Miller, and Herbert A. Hauptman
discusses their Shake-and-Bake method of molecular structure determination
as well as parallel implementations of the Shake-and-Bake method.
This very long 2-page spread discusses the phase problem of x-ray
crystallography as well as the groups Shake-and-Bake solution and parallel
computing implementation.
The article also provides an excellent explanation of the problem and history
of x-ray crystallography before discussing the members of the team
and their solution strategy.
Dr. Russ Miller
is quoted as stating that "we need an enormous amount of computational power
and the only way to get it is through massively parallel computing." He
goes on to say that in terms of the procedure "you essentially toss the atoms
into a shoebox with enough knowledge to make sure the structure makes chemical
sense." The computer calculates a set of phases corresponding to these
atomic positions and then randomly perturbs them to arrive at a lower
correlation value. Then a new set of atomic positions is calculated and the
method continues back and forth between atomic positions and atoms as many as
200 times for a single trial structure. When enough trial structures have
been shaked and baked, the right answer will emerge as a significantly low
value based on Hauptman's correlation value. Finally, the article concludes
by listing a set of successes that the team has had since since they
derived the current form of the procedure in 1992, including approximately
15 molecules ranging in size from 28 to 500 non-H atoms.
- September, 1993:
Dr. Russ Miller,
Dr. Herbert A. Hauptman, and Dr. Charles M. Weeks appear in
"High-Tech Shower International," by Science & Technology News, a technology TV channel in Japan through satellite. Miller explains how Shake-and-Bake
is used to solve difficult molecular structures and presents a demonstration
on an SGI workstation at the Medical Foundation of Buffalo.
- June 2, 1993: An article in Inside R & D entitled
"Solve Complex Molecular Structures Fast" discusses the software devised
by
Dr. Russ Miller
and colleagues to routinely solve structures up to 400 non-H atoms from
x-ray diffraction data. It states that the software takes only hours
to solve such structures, which used to take years to solve, and have been
run on a Thinking Machines Connection Machine CM-5 parallel supercomputer.
The article states that a key virtue of the software lies in its ability to
solve complex structures with virtually no input from the user, given data
with sufficient resolution. The article concludes by stating that there
are two versions of the program available, one for Unix workstations and
the other for the aforementioned CM-5.
- April/May 1993: An article in PSC News entitled
"Shake and Bake Parallel Solutions to the Phase Problem in X-Ray
Crystallography" displays a nice picture of
Dr. Russ Miller,
Nobel Laureate Herbert A. Hauptman, and Dr. Charles M. Weeks.
The two page article starts by describing the phase problem of x-ray
crystallography and the fact that this group used massively parallel computing
and their Shake-and-Bake algorithm to reduce the time for solving
molecular structures from years to hours. They state that Shake-and-Bake
addresses a problem that has help up the progress of molecular biology
for decades as x-ray crystallography is the primary tool for determining
molecular crystal structures. The article states that prior to the
Shake-and-Bake method, structures up to 100 non-H atoms could be solved
routinely by direct methods, but Shake-and-Bake has broken that barrier and
shows no limitations, except for the quality of the input data. The article
quotes Miller as saying that "we need an enormous amount of computational power
and the only way to get it is through massively parallel computing." Miller
goes on to describe the algorithm by saying that "you essentially toss the atoms
into a shoebox with enough knowledge to make sure the structure makes chemical
sense. The computer calculates a set of phases corresponding to these
atoms and then randomly perturbs them to arrive at a lower value of
Hauptman's minimal function. Then a new set of atomic positions is
determined from these phases and the cyclical process is repeated.
Once enough trials have gone through this shake and bake process, the right
answer will emerge by evaluating the resulting minimal function values.
Finally, the article states that since 1992, the results have been quite
encouraging. The team has used both a Connection Machine CM-2 and CM-5
for parallel computing development.
- April, 1993: An article appears in the President's Column of
Impact that discusses the Shake-and-Bake team's success with
solving Crambin, a molecule that consists
of more than 400 atoms. It states that the principle authors of
the program are Dr. Charles M. Weeks and
Dr. Russ Miller
and that Shake-and-Bake was used to solve Crambin
in a routine and automated fashion. The article goes
on to discuss the importance of this result and the recent funding received
by the NIH to fund this project. Herb Hauptman calls this result
"by far the most dramatic application of the new mathematical method
devised by this team.
- April 29, 1993: An article in the states that
Dr. Russ Miller
has accepted an invitation to join the program committee of the
12th International Conference on Pattern Recognition, which will be held in
October in Israel.
- April 1, 1993: An article in the Reporter states that
Dr. Russ Miller
has accepted an invitation to join the program committee for
IPPS'94, the Eighth International Parallel processing Symposium. It also states
that Miller has performed groundbreaking research in the area of parallel
algorithms and architectures, specifically applied to areas that include
image processing, computational geometry, graph theory, combinatorial optimization and computational crystallography. The article concludes with a short
professional biography of Miller.
- March 11, 1993: An article in the Reporter entitled
"Using massive computers, team first to solve protein structure" states
that after a group in Germany published a paper stating that they were
unable to unravel the molecular structure of a given protein,
Dr. Russ Miller
was able to use the Shake-and-Bake method developed by a research team
including himself and members of the Medical Foundation of Buffalo
to solve this large and difficult structure. It was the first time a
"direct methods" procedure was able to solve a structure of this size
and complexity. The solution to this protein structure is the latest
in a series of spectacular successes for this research group, according
to the article.
- February 19, 1993: An article in the
Buffalo News
entitled "Area researchers solve molecular mystery" states that
by using their Shake-and-Bake algorithm, the Shake-and-Bake team solved
a 400+ atom structure. The article quotes Dr. Russ Miller
as saying that "we were hoping to get one good solution; we were surprised when
we got 16." The article states that Miller developed new techniques
for some of the routines and adapted the program to run on massively parallel
supercomputers.
- February 18, 1993: A brief note in the Reporter
states that work by a team of researchers including
Dr. Russ Miller and
Nobel Laureate Herbert A. Hauptman was used to solve two unknown crystal
structures overnight and that these results were discusses in the
November issue of The Science Teacher. The note also states that
the story also appeared in Science News.
- February 4, 1993: An article in the Reporter
states that
Dr. Russ Miller
has been appointed ot the executive committee of the newly formed
Technical Committee on Parallel Processing of the Institute of
electrical and Electronics Engineers (IEEE).
It also goes on to define parallel computers and gives a brief biography on
Miller.
- January/February 1993: An article in EDUCOM, vol.2, no. 1,
entitled "New Method Unravels Molecular Structure" has a nice picture
of Dr. Russ Miller
and states that Dr. Miller has used the Shake-and-Bake program to solve
two previously unknown and difficult structures literally overnight.
The article states that this research will profoundly accelerate
scientific advances in rational drug design.
- January 13, 1993: An article in the Amherst Bee states
that Dr. Russ Miller
has been appointed to the executive committee of the newly-formed
Technical Committee on Parallel Processing of the Institute of Electrical and
Electronics Engineers and that his area of responsibility will be parallel
processing education.
- January 31, 1993: An article in the
Buffalo News states that
Dr. Russ Miller
has been appointed to the executive committee of the newly formed
Technical Committee on Parallel Processing of the Institute of
electrical and Electronics Engineers (IEEE).
It states that Miller's primary responsibility will be in the area of
parallel processing education. The article goes on to discuss
Miller's work in parallel computing, his work with members of the Medical
Foundation of Buffalo, and, in particular, his work with Nobel Laureate
Herb Hauptman on the Shake-and-Bake method of molecular structure
determination.
|
