Table 1.

A comparison of candidate gene and genome-wide approaches to genetic association studies

Candidate Gene	vs.	GWAS
A very small fraction of the genome is typically analyzed		Covers up to 80 % of the genome
Required a strong a priori hypothesis		Is a hypothesis-free or hypothesis-generating approach
A more statistically powerful approach		Compared to candidate genes, the needed correction for multiple testing leads to low power
Suitable for identifying variants of large and small effect		Typically more suitable for identifying variants of larger effecta
Can include rarer variants and nonSNP variants such as variable number of tandem repeats (VNTRs)		Traditionally includes common variation (SNP occurring in <1 % of the population, although some GWAS chips contain rarer variants and/or are customizable)

^aThis is dependent on sample size