Linkage and association

Genet Epidemiol. 1998;15(6):565-76. doi: 10.1002/(SICI)1098-2272(1998)15:6<565::AID-GEPI2>3.0.CO;2-J.


Methods of both linkage analysis and association analysis may be model-based or model-free. The former are useful for initial exploratory analysis, the latter for more detailed multivariate genometric analysis. Linkage leads to an association, but that association may be solely intrafamilial. Allelic association may be due to pleiotropy, linkage disequilibrium, meiotic drive, selection, or population stratification. Using non-transmitted parental alleles as controls for alleles transmitted to cases, in conjunction with a McNemar-type test, does not detect association in the absence of linkage. Model-based analyses should use models that approximate the complexity of the disease being studied in order to be both robust and powerful.

Publication types

  • Research Support, U.S. Gov't, P.H.S.
  • Review

MeSH terms

  • Case-Control Studies
  • Chromosome Mapping
  • Data Interpretation, Statistical*
  • Genetic Linkage / genetics*
  • Genotype
  • Humans
  • Likelihood Functions
  • Linkage Disequilibrium / genetics
  • Lod Score
  • Meiosis / genetics
  • Models, Genetic*
  • Multivariate Analysis*
  • Phenotype
  • Reproducibility of Results
  • Selection, Genetic
  • Statistics, Nonparametric