Existing guidelines for crystallizing proteins generally require subst
antial amounts of protein and scarce laboratory resources. To address
this problem we have developed the program DESIGN, which, suited to da
ily laboratory usage, utilizes statistics when screening variables of
crystallization, The program DESIGN is described in considerable detai
l and examples of its outcome and capabilities are shown. In addition,
a new scoring scale for crystallization trials is described. It is ex
pected that the use of such an approach may provide additional constra
ints toward crystallizing the protein, even if the conditions applied
are not successful. (C) 1994 Academic Press, Inc.