Abstract
摘要:生技製藥是台灣二十一世紀的關鍵工業,藥物基因體學與微陣列是過去十年最重要的科技突破,而且在疾病的偵測與治療上有非常大的潛能與應用。因病人的基因變異與基因-環境的交互作用,所以病人對臨床處置的反應亦有所不同。在人類基因體計畫完成後,疾病的分子目標可被鑑別,DNA遺傳標的、突變與表現形式亦可以利用微陣列技術製成生物晶片診斷試劑,而且在基因體時代進行個別化臨床處置的目標臨床試驗時,鑑定分子標的診斷試劑重要性日增,藥物與診斷試劑合併研發是達到個人化醫藥最基本的步驟,所以微陣列生物晶片診斷試劑之品質與目標臨床試驗設計、分析是非常重要的。因此本研究專題為三年計畫共包括三部分:(1) 微陣列生物晶片診斷試劑之基因篩選與所使用的基因複合生物標的分類指標(Genomic Composite Biomarker Classifier , GCB 分類指標)之最佳表示方式的研究, (2) 微陣列資料的一致性,再現性與系統性偏差之評估研究與(3)目標臨床試驗的療效與樣本數的估算研究。
對於第一部分,我們將決定差異表現基因之臨界值考慮在統計假說內,並導衍其統計檢定量。而對於GCB分類指標,我們將假設選
Abstract: Biotechnology and Biopharmaceutical industry is the key to Taiwan in the 21th century. Pharmacogenomics and microarrays are two of the most important scientific breakthroughs in the last decade and present great potentials in detection and treatment of diseases, and many other applications. On the other hand, due to genetic variations and genetic-by-environmental interaction, patients respond differently to the same treatment or therapeutic regimen. After completion of the Human Genome Project, the disease targets at the molecular level can be identified and hence biochip products based on heritable DNA markers, mutations, and expression patterns for detection of diseases using the microarray technology is possible. Therefore, the importance of diagnostic tests for identification of molecular targets increases as more targeted clinical trials will be conducted for the individualized treatment of patients in the genomic era. As a result, the drug-device co-development is the basis to the ultimate goal of personalized medicine. However, the quality of the microarray-based diagnostic biochip and innovative design and evaluation of efficacy for targeted clinical trials are vital to achieve the goal. Statistics plays an important and indispensable role in evaluation of the quality of diagnostic biochip and effectiveness of molecular-targeted drugs using the targeted clinical trials. Hence this 3-year research consists of three parts: (1) selection of the optimal representation for the genomic composite biomarker (GCB) classifier with differentially expressed genes using in the microarray-based diagnostic biochips, (2) assessment of agreement, reproducibility, and systematic bias of microarray data, and (3) evaluation of effectiveness and sample size estimation for targeted clinical trials. For the first part, the hypothesis will be formulated by taking into consideration the cut-off values for identification of differentially expressed genes. A statistical procedure will be developed with respect to the hypothesis. For the GCP classifiers, under the assumption that the number of genes selected in the GCP classifiers is also a random variable, the distribution of the GCP classifiers will be derived using the compound Poisson distribution. In addition, the procedure for selection of the optimal form for the GCP classifiers will be developed along with the method for determination of its thresholds. For the second part, confidence limits of the concordance correlation coefficient (CCC) and intraclass correlation coefficient (ICC) will be suggested for evaluation of agreement and reproducibility respectively. Based on the concept of generalized pivotal quantities (GPQ), the exact confidence intervals will be derived for CCC and ICC due to the small sample sizes of microarray experiments. We will develop a procedure for estimation of systematic bias when both measurements are subject to errors and expression levels are correlated among genes. For the third
Keyword(s)
藥物基因體學
微陣列
基因複合生物標的分類指標
一致性
再現性
指標臨床試驗
Pharmacogenomics
Microarray
Genomic Composite Biomarker Classifier
Agreement
Reproducibility
Targeted Clinical Trials