Abstract:
It is now possible to use complete genetic linkage maps to locate major
quantitative trait loci (QTLs) on chromosome regions. The current methods
of QTL mapping (e.g., interval mapping, which uses a pair or two pairs
of flanking markers at a time for mapping) can be subject to the effects
of other linked QTLs on a chromosome because the genetic background is
not controlled. As a result, mapping of QTLs can be biased, and the resolution
of mapping is not very high. Ideally when we test a marker interval for
a QTL, we would like our test statistic to be independent of the effects
of possible QTLs at other regions of the chromosome so that the effects
of QTLs can be separated. This test statistic can be constructed by using
a- pair of markers to locate the testing position and at the same time
using other markers to control the genetic background through a multiple
regression analysis. Theory is developed in this paper to explore the idea
of a conditional test via multiple regression analysis. Various properties
of multiple regression analysis in relation to QTL mapping are examined.
Theoretical analysis indicates that it is advantageous to construct such
a testing procedure for mapping QTLs and that such a test can potentially
increase the precision of QTL mapping substantially.