EffiSim version 1.1
Last modified 29th of March 2006
The main change in this version is that it now considers regional screens
as well as genome-wide screens. The user must now specify either that the
screen in question is genome-wide screen (as 'g') or provide the genetic
length of the target region (in cM).
The measure of efficiciency that incorporates weights is now the 'cost per
mutation' (or CM), rather than its reciprocal, the number of mutations
screened per unit cost (or work).
The number of mutations inherited per G1 is now a random variable. It follows
the Poisson distribution with a mean specified using the '-x' option. This
allows us to consider regional screens where the number of mutations we expect
is not an integer.
I spotted a few minor bugs around the place, most of which were new and occured
while making the changes specified above.
EffiSim version 1.0
Last modified 19th of December 2005
We have shown that estimates for the variability which were derived under the
assumption of no linkage are biased and will provide underestimates. For this
reason, EffiSim only provides estimates of the variability of the number of
the number of mutations screened by simulation.
The theory for the efficiency calculations has been properly developed for the
y = 1 case. Although theoretical estimates for the y > 1 case are provided, we
recommend that they be verified using the simulation option.
For any questions/bug reports/comments, please contact: silver@wehi.edu.au