Simple power calculation
300 Words | Approximately 2 Minutes Read | Last Modified on August 12, 2014
Power of Cochran-Armitage Trend Test for Case Control Association Studies
This is an implementation of https://doi.org/10.1159/000064976. Calculation of $\hat{\sigma}_a$ implements the matrix notation of https://doi.org/10.1002/gepi.10245 (unfortunately this paper is severely flawed with numerous typos in maths)
Latest Version
Usage
catt -h
usage: catt [-h] [-b f0] [-r gamma] [-p MAF] [-R N] [-S N]
[-m {additive,dominant,recessive}] [-a ALPHA]
[--alternative {two-sided}]
Cochran-Armitage trend test (CATT) power calculation for case/ctrl genetic
association studies, implementing Freidlin et al, Hum Hered
2002;53:146-152
optional arguments:
-h, --help show this help message and exit
-b f0, --baseline_penetrance f0
wildtype genotype penetrance, default to 0.01
-r gamma, --odds_ratio gamma
odds ratio, default to 1.0
-p MAF, --maf MAF minor allele frequency, default to 0.1
-R N, --num_cases N number of cases, default to 500
-S N, --num_ctrls N number of ctrls, default to 500
-m {additive,dominant,recessive,multiplicative}, --moi {additive,dominant,recessive,multiplicative}
mode of inheritance, default to "additive"
-a ALPHA, --alpha ALPHA
significance level (test size), default to 0.05
--alternative {two-sided}
alternative hypothesis, default to two-sided
Example
This bash script demonstrates the use of the program. Run it to have a taste!
maf="0.1 0.2 0.3"
OR="1.1 1.2 1.3 1.4"
N="200 500 800"
f0=0.01
moi='additive'
alpha=0.05
for i in $maf; do
for j in $OR; do
for k in $N; do
power=`catt -b $f0 -r $j -p $i -R $k -S $k -m $moi -a $alpha 2> /dev/null`
echo $power $i $j $k $f0 $alpha $moi
done
done
done
Output of the program is disease prevalence (written to stderr stream) and power (written to stdout) stream. To display power only, you can redirect the stderr stream by:
catt ... 2> /dev/null