Traditional approaches for genetic mapping are to simply associate the genotypes of a quantitative trait locus (QTL) with the phenotypic variation of a complex trait. Amoremechanistic strategy has emerged to dissect the traitphenotype into its structural components andmap specific QTLs that control themechanistic and structural formation of a complex trait.We describe and assess such a strategy, called structuralmapping, by integrating the internal structural basis of trait formationinto aQTLmapping framework.Electrical impedance spectroscopy (EIS) hasbeeninstrumental for describing the structural components of a phenotypic trait and their interactions.By building robustmathematical models on circuit EIS data and embedding these models within a mixture model-based likelihood for QTL mapping, structuralmapping implements the EM algorithm to obtain maximum likelihood estimates of QTL genotype-specific EIS parameters.The uniqueness of structuralmapping is tomake it possible to test a number of hypotheses about the pattern of the genetic control of structural components.We validated structuralmapping by analyzing an EIS data collected forQTLmapping of frost hardiness in a controlled cross of jujube trees.The statistical properties of parameter estimates were examined by simulation studies. Structuralmapping can be a powerful alternative for geneticmapping of complex traits by taking account into the biological and physicalmechanisms underlying their formation.
All Science Journal Classification (ASJC) codes
- Information Systems
- Molecular Biology