We are aware that most of the genetic association studies are under-powered just because of the fact the sample size is inadequate. The major reason for the power to be small is because, in most situations either or both the risk allele frequency and the prevalence of the phenotype tend to be very small. This demand a very high sample size (in thousands) to either accept or reject null hypothesis (though power matters more in negative studies which fail to reject null hypothesis). But there are situations when the risk allele frequency (0.5-0.7) and prevalence of phenotype are high (50-70 %), in which case the sample size that is required to show a decent effect size comes down to range of few tens (20-40). However, the journals either reject the submission or recommend us to mention that sample size is small. I don't know why. Comments regarding this are welcome.
If you haven't already done so, I think the reviewers would probably like an explanation of the assumptions used in your power calculation.
CAD has quite a broad definition so estimates vary, but ~20% would be a reasonable assumption in an adult population, I would look for a reference on prevalence in a population similar to yours. In a power calculation your 'effect allele' frequency can never be greater than 0.5, even if the allele contributing to disease is the major allele. This could also apply to disease prevalence, if greater than 50% you're actually investigating a protective effect, regardless of how the reults are presented.
@Jennifer Bolton. Thank you for that reply. Can you elaborate on power calculation for an allele with protective effect for a phenotype. In that case prevalence should be that of normal population or diseased population? How do I mention the effect size? Because the odds ratio will be less than one, right? I am unclear with this.
I am not familiar with coronary disease literature but one major reason for small effect in psychiatric disorders is hererogeniety. Any single locus contribute a small fraction of the risk.
how to calculate sample size for initiating a study after knowing minor allele frequency and odds ratio, for a gene in case control study.
In most of the research articles authors will give odds ratio and genotype frequencies, in few articles they wont give odds ratio at all, I have selected few studies and calculated minor allele frequencies for that
for example : Haptoglobin gene
GF MAF controls
Hp 1-1 0.408 985
Hp 1-2 0.75
Hp 2-2 0.59
Hp 1-1 0.35 1079
Hp 1-2 0.64
Hp 2-2 0.64
sample size for this study is
recessive model 471
log additive model 217
dominant model 814
my query is does my calculation is correct or not I have done it with quantos software
which model should I choose recessive or dominant or additive
additive is showing less sample size should I prefer it for initiating a study
please help me out
to do power analysis to estimate your sample size, you have to write your hypothesis, and based on that you decide what statistical test you will use. It should be one of the inferential statistics. so you need to determine the following: alpha {standard to be .05}, power [standard to be .80], effect size {small, moderate, or large, each test has its own value, you can find these values in the net}. Then download free programs to calculate the sample size such as G. power.
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping