R. Agarwala et al., Inverse inbreeding coefficient problems with an application to linkage analysis of recessive diseases in inbred populations, DISCR APP M, 104(1-3), 2000, pp. 3-44
Medical geneticists connect relatives having the same disease into a family
structure called a pedigree. Genetic linkage analysis uses pedigrees to fi
nd the approximate chromosomal locations of disease-causing genes. The prob
lem of choosing a pedigree is particularly interesting for diseases inherit
ed in an autosomal recessive pattern in inbred populations because there ar
e many possible paths of inheritance to choose from. A variety of shortcuts
are taken to produce plausible pedigrees from inbred populations. We lay t
he mathematical foundations for a shortcut that was recently used in a pedi
gree-disease study of an inbred Mennonite population. Recessive disease gen
es can be localized using the shortcut of homozygosity mapping by finding r
egions of the genome where affected persons are homozygous. An important qu
antity in homozygosity mapping is the inbreeding coefficient of a person, w
hich is the prior probability that the person inherited the same piece of D
NA on both copies of the chromosome from a single ancestor. Software packag
es are ill-suited to handle large pedigrees with many inbreeding loops. The
refore, we consider the problem of generating small pedigrees that match th
e inbreeding coefficient of one or more affected persons in the larger pedi
gree. We call such a problem an inverse inbreeding coefficient problem, We
focus on the case where there is one sibship with one or more affected pers
ons, and consider the problem of constructing a pedigree so that it is "sim
pler" and gives the sibship a specified inbreeding coefficient. First, we g
ive a construction that yields small pedigrees for any inbreeding coefficie
nt. Second, we add the constraint that ancestor-descendant matings are not
allowed, and we give another more complicated construction to match any inb
reeding coefficient. Third, we show some examples of how to use the one-sib
ship construction to do pedigree replacement on real pedigrees with multipl
e affected sibships. Fourth, we give a different construction to match the
inbreeding coefficient of one sibship, while attempting to minimize a measu
re of the inbreeding loop complexity. (C) 2000 Elsevier Science B.V. All ri
ghts reserved.