The development and dissemination of this software is supported, in part, by grants EY12565 from the National Institutes of Health.


QTDT is a package written in C++ which performs linkage and association analyses. The main contribution of QTDT is an implementation of the Fulker et al (1999) family-based test of association using a variance components framework. QTDT will perform the Fulker et al (1999) test, for an arbitrary number of markers and alleles, in general pedigrees (Abecasis et al, 2000a, 2000b).


MERLIN is the acronym for Multipoint Engine for Rapid Likelihood Inference that is currently under development (Abecasis et al, 2002). The program uses sparse binary trees to represent patterns of gene flow in general pedigrees and uses a modification of the Idury and Elston (1997) algorithm for full multipoint calculations --- haplotyping, error-checking, IBD, and kinship --- with considerable gains in speed and memory efficiency over other programs. This feature will make it especially useful for analyses using dense SNP maps.

Conditioning on traits

We distribute a SAS program to perform variance components linkage analysis of selected and non-normal sib-pair samples (Sham et al, 2000b). This has been re-implemented in a C++ program to facilitate further development.


This SAS macro package performs DeFries-Fulker and Haseman-Elston linkage analysis in selected or random samples. The program outputs for each position at which an estimate of IBD is available a t statistic, associated p value, and optional outlier influence statistics. Multipoint linkage plots are also available. As with the two other packages listed above, QMS2 utilizes IBD calculations obtained from Genehunter or MERLIN. We will continue to incorporate all our regression-based methods into this package.


A Windows-based application for choosing those individuals for genotyping, from an unselected phenotyped population, who will maximize power for linkage (Purcell et al, 2001).

Genetic Power Calculator

A website has been constructed to perform the linkage and association power calculations presented in Sham et al(2000a), based on our variance components framework. It is hoped that this website will satisfy the needs of most researchers in obtaining the basic power calculations necessary before conducting a genetic study. It also includes an online CGI interface to a C++ program used for the methodological work on epistasis.


Graphical Representation of Relationships. This program computes the mean and standard deviation of allele sharing in state across the genome for all pairs of individuals in a sample and graphically displays this information for the purpose of detecting pedigree errors. For typical genome scan marker sets, there is more than enough information to unambiguously determine if sibs, half-sibs, MZ twins, and parent-offspring pairs are related as believed, and GRR makes detection of such errors a simple and quick process (Abecasis et al,2001)


Page last modified 08/26/08
contact webmaster