MARKOV CHAIN MONTE CARLO FOR VERY LARGE PEDIGREES

我们建议设计、测试和实施新的方法来估计癌症易感性综合征的携带者概率、外显率和等位基因频率。我们将使用马尔可夫链蒙特卡罗（MCMC）方法和相关的样本路径方法。家谱和疾病数据将从犹他州人口数据库（UPDB）获得，该数据库由100多万与癌症相关的基于人口的家谱记录和随访数据组成。我们将首先模拟系谱成员之间的疾病发生，并尝试用MCMC方法估计相关的遗传参数。一旦我们成功地估计了载子概率，我们将扩展方法来估计外显率和等位基因频率。当MCMC算法能够正确估计所有所需的参数时，我们将使用它来使用乳腺癌、前列腺癌、结直肠癌和卵巢癌的实际UPDB数据来生成载体概率、外显率和等位基因频率的估计。

We propose to design, test, and implement new methods for estimating carrier probabilities, penetrance, and allele frequency for cancer predisposition syndromes. We will use Markov Chain Monte Carlo (MCMC) methods and related sample-path methods. Pedigree and disease data will be obtained from the Utah Population Database (UPDB), which consists of more than one million population-based genealogical records linked to cancer and follow-up data. We will begin by simulating the occurrence of disease among members of the pedigrees and attempting to estimate the relevant genetic parameters with MCMC methods. Once we have successfully estimate carrier probabilities, we will extend the methods to estimate penetrance and allele frequency. When the MCMC algorithm is able to correctly estimate all the desired parameters, we will use it to generate estimates of carrier probabilities, penetrance, and allele frequency using actual UPDB data on breast, prostate, colorectal, and ovarian cancer.