E0065
Gearing Up for Structural Genomics: The Challenge of
Hundreds of Proteins and Hundreds of Thousands of Crystallization Experiments
Per Year. G.T. DeTitta1, J.R. Luft1, J.
Wolfley1, R. Collins1, M. Bianca1, D.
Weeks1, I. Jurisica2, P. Rogers2, J.
Glasgow3, S. Fortier3, 1Hauptman-Woodward
Inst., 73 High St,. Buffalo, NY 14203, 2Univ. of Toronto, 140 Saint
George St., Toronto, Ont. M5S 3G6 Canada, 3Queen’s Univ.,
Kingston, Ont. K7L 3N6 Canada
Structural genomics promises to yield hundreds of proteins
each year for structural analysis. The challenge to crystal growers is to keep
pace. Our approach is to combine high throughput (HTP) crystallization setup
and evaluation with sophisticated algorithmic analyses of the outcomes for the
purposes of recipe prediction.
In the wet lab we have the capacity to prepare and evaluate
the results of over sixty thousand (61.4K) crystallization experiments a
workweek. Each is a microbatch experiment conducted under paraffin oil.
Experiments are held in 1536-well micro-assay plates, each well of which
contains a chemically distinct crystallization cocktail. Robotic pipetting
allows the deployment of 200 nanoL droplets of protein stock to each of the
wells of a plate in less than five minutes, allowing us to handle unstable
proteins. Current total protein requirements are likely to be in the 10 mg
range. After setup, plates are placed on a computer controlled XY table and
translated under a digital camera where images are captured. The XY table can
accommodate 28 plates (43K experiments) and the camera can record 43K images in
approximately twelve hours.
Images are analyzed automatically to determine the outcomes
of the crystallization experiments. We are developing a standard vocabulary of
outcomes that describe the results: clear drop, amorphous precipitate, phase
separation, microcrystals, crystals, and uncertain outcome. These outcomes,
recorded as a function of time, are the cornerstone of a crystallization
database that will contain physical information about individual proteins and
results of crystallization experiments. Using case-based reasoning and data
mining algorithms we will identify patterns of similar properties and
crystallization outcomes relating two or more proteins in the database. Our
hypothesis is that, given a quantitative measure of “similarity”
between proteins, recipes successfully employed for one protein will be useful
starting points for crystallization experiments with similar proteins. Future
work will center upon the most predictive measures of
“similarity”.
Work supported in the United States by the John R. Oishei
Foundation and NASA NAG8-1594 and in Canada by NSERC and CITO.