Genetic risk prediction: individualized variability in susceptibility to toxicants

Genetic risk prediction uses genetic data to individualize prediction of outcome or effect from a known harmful toxicant. Several examples of toxicogenetics (usually binary traits) are discussed, reflecting largely Mendelian traits before the Human Genome Project began in 1990. Numerous complexities of the genome and what constitutes "a gene" have emerged during these past two decades. Examples of toxicogenomics (continuous outcomes, gradients) are examined. Most xenobiotic-induced environmental diseases resemble human complex diseases or other multifactorial traits such as height; these traits result from hundreds of low-effect genes. Consequently, uncovering an association between a trait and a genetic variant in a large cohort can provide important information about underlying biology; however, screening for a specific variant in an individual worker or patient has poor predictive value and little clinical utility. Individualized risk assessment for toxicants that cause environmental diseases, although a lofty goal, remains to be achieved.