In this Perspective, we evaluate the extent of rare coding variation in empirical data, discuss data processing and quality control of raw sequence data, review analytical methods for detecting genotype-phenotype associations, their expected statistical power, and the potential for confounding due to population stratification. To illustrate our arguments, we used empirical whole-exome sequence data from 184 individuals from the International HIV Controllers Study40 and 254 control individuals from Schizophrenia (SCZ) exome sequencing study.
