致癌基因樹的重建

Abstract

比較型基因組雜交分析技術 (CGH)，是一種分子生物學上測量腫瘤細胞中，染色體複製倍數(增生或減少)的方法。在生物領域中，通常樹模型是用於研究生物的演化、物種間的進化關係。近幾十年來，生物學家認為癌症發生的過程和DNA的增生或減少息息有關，而且不單單只是某一個基因改變，而是一連串不同基因改變事件發生。因此，也將樹模型當作分析CGH資料的工具，用來研究、探索癌症的發展過程(carcinogenesis)。在數學的基礎下，我們可以從CGH的資料去推導出數學模型。 Desper 及其作者提出了分支樹和距離樹兩種模型，比 Vogelstein 等學者提出的直腸癌的單路徑模型更加精確。在這篇論文中，我們介紹了分支樹和距離樹兩個不同的樹模型，比較兩者的相異處。另外，也利用最大概似估計法，來建立引起腫瘤、癌症的樹模型(oncogenetic tree model)。

Comparative Genomic Hybridization (CGH) is a molecular cytogenetic method to measure the copy number aberrations of chromosomal in tumor cells. In biological domain, tree model is usually used on the evolutionary relationship between organic and different biological species. In decades, the biologists believe that increase and decrease of DNA cause the occurrence of cancer. Furthermore, not only the change of one genetic event results in the cancer occurrence, but also serial alteration of different genes. Therefore, tree model is used as the tool to analyze CGH data and explore carcinogenesis. The mathematical model from CGH data can be inferred under mathematical ground. Desper et al. stated two models: maximum-weight branching model and distance-based tree model. These two models are more accurate than path model of colorectal-tumor which was brought up by Vogelstein et al.. In this paper, we introduce two different tree models and compare the differences between them. In addition, the oncogenetic tree model is reconstructed by using the idea of maximum likelihood.