Statistical models for genetics data are often surprisingly challenging, and often require advanced and new statistical methods. This project continues to investigate a number of such areas, including: extension of a scheme for estimation of linkage from transmission/disequilibrium data (see [3]), and a new comprehensive approach to genome scans (see [5]). One study looked at the distribution of genes on the Drosophilia X chromosome (see [1]), while another looked at the spacings between mouse genes and retrovirial tags as a method for detecting new oncogenes (see [7]). The allele transmission study [2,3] is the first to estimate TDT linkage and association model parameters from data collected jointly from both parents and all affected children. Using a series of realistic simulations we show that our TDT estimation methods are essentially as powerful and robust as the classical TDT, while our confidence intervals for linkage and association provide significantly new useful information. All results were thorougly tested using a series of realistic simulations and models, and demonstrated that our estimation methods are a valuable contribution to practical genetics data analysis. We have recently extended the approach to sampling plans that use allele transmision data from both parents. Simulations confirm that this approach gives increased power to detecting (separately) association and linkage. The genome scan technology (see [4,5,6]) involves the use of importance sampling to accurately estimate the level of significance for multiple testing using many markers, several thousand at a time. It is shown to be significantly more efficient than naive Monte Carlo testing, and handles the case of unequally spaced markers. Previous work in this area used large deivation probabilistic methods that did not address the case of unequally spaced markers. Also, the importance sampling approach extends to multiple testing in Hardy-Weinberg problems, multiple allele TDT testing, and many other mutiple testing problems in genetics. [1] Parisi, Nuttall, et al., Paucity of genes on the Drosophilia X chromosome showing male-biased expression, Science, 299: 697-700 (2003), [2] J.E.Bailey-Wilson, J.D.Malley et al.: Comparison of Novel and Existing Methods for Detection of Linkage Disequilibrium Using Parent-Child Trios in GAW 12 Genetic Isolate Simulated Data. Genetic Epidemiology, 21(Suppl 1): S378-S383 (2001). [3] Estimation of Linkage and Association from Allele Transmission Data. Biometrical Journal, 45: 349-366 (2003). [4] Results were presented as an invited talk at the Intl. Gen. Epi. Soc. meeting in Garmisch-Partenkirschen, Germany (Sept. 2001). [5] Malley, Naiman and Bailey-Wilson (2002) Comprehensive Method for Genome Scanning. Human Heredity, 54: 174-85 (2002). [6] Klein, Kovac, et al., Importance sampling method for correction for multiple testing in affected sib-pair linkage analysis. Genetic Analysis Workshop 8, IGES meeting, New Orleans (Nov. 2002). To appear, Genetic Epidemiology Supplement, 2003. [7] Suzuki, Shen, et al., New genes involved in cancer identified by retroviral tagging. Nature Genetics, 32: 166-174 (2002).
Showing the most recent 10 out of 24 publications
