Health Research and Policy

Statistical Software


Documentation and software program


Documentation and Family Genotype Analysis Program. This program can be used to detect association between a disease and a diallelic polymorphism of a candidate gene by analyzing genotypes in arbitrary types of nuclear families.

FGAP is written in Java and hence must be run from a computer on which there is a Java runtime environment. The program can be run using a Graphical User's Interface, from the command line, or in "batch." See the associated "README" file for full details.

Two likelihood-based score statistics can be computed. The first statistic, the nonfounder statistic (NFS), extends the transmission disequilibrium test to accommodate affected and unaffected offspring and missing parental genotypes. The second statistic, the founder statistic (FS), compares observed or inferred parental genotypes to those of some reference population. In this comparison, the genotypes of affected parents or those with many affected offspring are weighted more heavily than unaffected parents or those with few affected offspring. Genotypes of single, unrelated cases and controls can be included in the analysis.

A detailed description of the statistical underpinnings of the software can be found in the following articles:

Shih M-C, Whittemore AS. Tests for genetic association using family data. Genet Epidemiol 22: 128-145, 2002.

Whittemore AS, Tu I-P. Detecting disease genes using family data. I. Likelihood-based theory. Am J Hum Genet 66: 1328-1340, 2000.

Tu I-P, Balise RR, Whittemore AS. Detecting disease genes using family data. II. Application to nuclear families. Am J Hum Genet 66: 1341-1350, 2000.

Clayton D. A generalization of the transmission/disequilibrium test for uncertain-haplotype transmission. Am J Hum Genet 65: 1170-1177, 1999.


retrovar computes the asymptotically correct variances for the estimates of the parameters from a logistic regression done with data which have been ascertained retrospectively and have certain covariance structures. See, "Logistic Regression of Family Data From Retrospective Study Designs"; A. S. Whittemore and J. Halpern; Genetic Epidemiology 25:177-189 (2003).


weightedKAC - description to be supplied soon.

Footer Links: