A major project of this section is the development of new statistical genetics methodology as prompted by the needs of our applied studies and the testing and comparison of novel and existing statistical methods. ? ? This year, we have continued our collaboration with Dr. Silvano Presciuttini and Dr. Fabio Marroni from the University of Pisa, on development of better methods for predicting mutation carrier status for known cancer genes in Italian families. We have extended the model development that we did for BRCA1 carrier status to develop models that allow improved prediction of MSH2 and MLH1 carriers. A paper presenting this method was published this year but is included in this section?s Genetic Epidemiology of Cancer annual report since it is a new method applied to colon cancer families from Italy.? ? The project to develop propensity scores in linkage analyses as a method for inclusion of covariate effects has been continued in conjunction with Betty Doan and Yin Yao. This method appears promising in that it is generally more powerful than including the covariates directly into the model, and does not have strongly inflated Type I error rates. Two papers have been published this year describing and evaluating this method [1, 2].? ? In addition, we have pursued another project designed to examine the effects of important environmental covariates on power and Type I error in linkage studies. We are simulating traits for which moderate to strong environmentsl risk factors play a role in risk of a disease trait (affected vs. unaffected) and then comparing the performance of various analytic methods that ignore covariates to the performance of methods that incorporate these covariates into the analysis. A manuscript is in preparation describing these results.? ? In this fiscal year we also devoted a large amount of time to evaluating existing methods for linkage, association and haplotype analysis using both STRP and dense SNP maps. We examined the effect of low to moderate LD on the Type I error rates of linkage statistics, and compared various methods of picking tagSNPs and building haplotypes in family data. Three original papers [3-5] were published as a result of this work, as well as two summary/review papers [6, 7]. ? ? We also used computer simulation techniques to determine the power and robustness of several types of linkage analysis with and without the inclusion of laboratory-determined (direct) haplotypes on some or all family members. We found that large increases in power can be obtained by including such information in the analysis, particularly when some family members are ungenotyped. A paper was published presenting these results [8].? ? We performed simulation studies to determine the effects on power and Type I error of utilizing molecular haplotyping techniques in linkage studies with various levels of missing genotype data. We found that these techniques do not inflate Type I error and substantially increase power when founder genotypes are missing. Importantly, most of the power gain can be obtained by only molecularly haplotyping 1 person per pedigree in most cases, a major savings in laboratory costs. A paper presenting these results has been published this year [9].? ? A review of the application of genetic methodologies to forensic science in the DNA victim identification process following the 9/11 World Trade Center disaster was published this year [10], reflecting the experiences of the advisory panel to the Office of the Medical Examiner, on which Dr. Bailey-Wilson was a member.? ? A review paper concerning pros and cons of cohort studies for detecting gene-gene and gene-environment interactions was published this year [11], incorporating results from a large number of power studies performed by this section for this purpose. ?
Showing the most recent 10 out of 21 publications
