Human genetic susceptibility to leprosy: methodological issues in a linkage analysis of extended pedigrees from Karonga district, Malawi
Thesis submitted 2003.
Leprosy is a disease of humans which is caused by infection with Mycobacterium leprae. Although infection with M. leprae is necessary for disease, it
is thought only a small proportion of those infected (probably less than 10%) develop clinical disease, which may be manifested across a wide spectrum.
Susceptibility to leprosy is influenced by both genetic and non-genetic factors, and there is evidence that genetic influences vary between populations.
Linkage analysis is a method for finding genes that influence a particular trait. Nonparametric methods of linkage analysis compare the identity by descent (IBD) sharing of genes among affected relatives to that expected in the absence of linkage. Often nonparametric linkage analysis is applied to
small families with multiple affected siblings, but extended multicase pedigrees may offer increased power to detect genetic determinants of disease.
This thesis deals with methodological issues raised in a linkage analysis of extended pedigrees with multiple cases of leprosy.
Power to detect linkage using affected relative pair data can be expressed as a function of an epidemiological parameter called the relative recurrence risk ratio (λ R ) which is defined as the ratio of the risk of disease in particular relatives of cases to the population risk of disease. It will be shown that power to detect linkage using relative trios can also be expressed as a function
λ R and a second, related parameter, λ R,R , defined here as the ratio of the risk of disease in individuals who have two affected relatives to the population
risk of disease. Estimates of λ R can be inflated if environmental risk factors are not properly accounted for. Methods for estimating λ R and λ RR while accounting for environmental risk factors are investigated using a marginal model, and estimates of λ R are presented for first, second and third degree relatives of people affected by leprosy in Karonga district, Malawi.
Further work examines the selection of members from extended pedigrees for inclusion in a linkage analysis. Simulation techniques are used to select
members in order to maximise information about IBD sharing between affected pedigree members without typing unnecessary members, which can be time-consuming and costly. Finally, nuclear families and extended pedigrees from Karonga are used in a partial genome screen to search for chromosomal
regions which may be linked to susceptibility either to leprosy per se or to leprosy type. Preliminary results are presented: no regions show significant
evidence of linkage according to the stringent criteria applied in genome screens, but there are regions that show evidence for potential linkage that
would be of interest to follow-up in this population.