The purpose of this project is to locate the genes leading to non-insulin- dependent diabetes mellitus (NIDDM). Recent enhancements to """"""""model free"""""""" linkage approaches coupled with molecular advances make a complete search of the human genome practicable and the likelihood for success high. To facilitate this search we have entered into a collaborative effort with Drs. Graeme Bell, Patrick Concannon, and Richard Spielman to make a 10 cM search of the entire genome. Dr. Bell will cover chromosomes 1A, Concannon 5-Il and Spielman 18-22 and the sex chromosomes. We will cover chromosomes 12 through 17 by typing 95 highly polymorphic markers (PIC equal to or greater than .70 and including known candidate loci) and testing for linkage. Three data resources will be employed. First, samples from 300 Mexican-American sib pairs from Starr County, Texas, where both are affected with NIDDM will be available. Second, DNA from 300 affected sib pairs will be provided to us by Dr. Bell. This sample will consist of approximately 180 Black pairs and 120 non-Hispanic White pairs. Lastly, we will use the 200 pedigrees (1,200 individuals) being collected by the American Diabetes Association (ADA). These samples should be available in the 3rd year of this project and will permit both non-parametric and parametric approaches to linkage. Use of the ADA resource will be enhanced because we are one of the family acquisition centers and the central data center and the University of Chicago is also one of the family centers. PCR will be performed using end-labelled oligonucleotides and variation detected via autoradiography of acrylamide sequencing gels. Linkage will be assessed by comparing the observed and expected identity by state (IBS) distribution for each marker using a chi-square statistic. Estimates from the literature suggest that a sample of 300 affected sib pairs should have sufficient power under a variety of genetic models. To further assess this issue, simulation studies were conducted assuming a 3 locus additive threshold genetic model for NIDDM. This model is on of the most difficult in which to detect linkage and actually represents a heterogeneity model. A marker linked to one of the loci was generated having 8 alleles and a PIC value of 0.81. Addition of the other 300 affected sib pairs pushes the power to 100%. Actual power will be bracketed by these values due to heterogeneity. Demonstration of linkage in more than l ethnic group, however, will be strong evidence that linkage is real. Successful identification of genes linked to NIDDM will provide a basis for understanding the etiology of the disease and its aggregation and will lead to the development of methods for reversing or slowing the progression from health to NIDDM and its complications.
Showing the most recent 10 out of 13 publications
