This is a response surface method experiment to find the treatment
conditions that optimize the yield of monoclonal antibodies produced
in mice innoculated with hybridoma cells. The response is proportional
to the number of antibody molecules produced. The experimental design
is a central composite design which has 2 experimental
factors each at five levels. The design consists of 11 runs and
includes 3 center points.
See Chapter 2 of Haaland (1989) for a more
complete description of this experiment. See also the help file for
abscrn.df
for a description of the screening experiment that
preceded this optimization experiment.
ARGUMENTS:
RadDos
an experimental factor giving the radiation dose in rads.
Prime1
an experimental factors giving the time in days between the
initial injection of Pristane
oil and the innoculation with antibody producing cells.
TtrVol
the response variable, antibody titer adjusted for volume. The
measured value
is proportional to the number of monoclonal antibody molecules
produced. The response should be maximized.
REFERENCES:
Haaland, P. D. (1989).
Experimental Design in Biotechnology.
New York: Marcel Dekker, Chapter 2.
SOURCE:
Used by permission of Marcel Dekker, Inc.
SEE ALSO:
EXAMPLES:
# This design is already available in S-PLUS under
# the name abrsm.df. These are the commands, that were
# used to create it.
abrsm.design <- rsm.design(2,factor.names =
list(RadDos=c(100,300),Prime1=c(7,21)))
abrsm.ttrvol <- c(207,257,306,570,100,513,315,
154,630,528,609)
abrsm.df <- cbind(abrsm.design,TtrVol=abrsm.ttrvol)
# Sample analysis
abrsm.rsm <- rsm.lm(abrsm.df)
summary(abrsm.rsm)
pareto(abrsm.rsm)
contour(abrsm.rsm,levels=c(100,200,300,400,500,
550,590,610))
surface(abrsm.rsm)
optim(abrsm.rsm)