|
The traditional boundaries between discovery, research, clinical development, therapeutic outcomes and consumer health data are being blurred as the need to integrate these data sources increases in importance. This shift is likely to lead to the convergence of entire industries, notably biotechnology, pharmaceuticals, consumer health care and computing - driving the development of new systems to manage this burgeoning supply of biomedical information.
Microarray methodology holds exciting promise for accelerating our understanding of disease, toxicology and therapeutic processes on the molecular level. Yet discovery organizations face a great challenge in developing the computational and statistical framework to analyze the massive amounts of data generated by microarray research.
The Comprehensive Cancer Trials unit (CCTU), which is under the Department of Clinical Oncology of the Chinese University of Hong Kong, has undergone a genomic research project on a Hepatocellular Carcinoma (HCC) (liver cancer) study. SAS has supported the Unit by providing SAS analytical software in the development of microarray analysis on the PC cluster platform. In addition, with the SAS technology, CCTU would be able to process the research work in simulation in a PC cluster set up with SAS clustering facility to speed up the computing process. As a result, a speedy microarray analysis on biomedical data could be carried out.
"This project will last until 2005 with the objective to develop the prognostic models in liver cancer using both clinical data and gene expression data from DNA microarrays," stated by Prof. Benny Zee, Professor/Director, Comprehensive Cancer Trials Unit, Department of Clinical Oncology of the Chinese University of Hong Kong. In a nutshell, the goal is to find out which or any type of gene expression can help identify which patient can have a better chance to live longer or improve the prognosis of patients in liver cancer. "The long-term significance is that the methodology developed in this study may apply to the analysis of DNA mcroarray data in other clinical settings and contribute to the area of quantitative bio-informatics research," Prof. Zee pointed. For this project, the data for analysis include DNA microarray data, Comparative Genomic Hybridization (CGH) Gain/Loss information at the chromosomes level as well as clinical data on patients.
Microarray experiments generate huge and complex multivariate data sets. The key challenges lie not only in generating these data but also in the development of computational and statistical tools to analyze the large volumes of data. Recent development of DNA microarrays technology has enabled a complete quantification of the whole genome (about 20,000 to 30,000 fields). A set of genes may encode proteins required for a particular function and contribute to the analysis of the clinical question. However, very few studies have taken into account the multivariate nature and potential interaction among the genes during prognostic modeling. "One of the difficulties is that the number of variables generated from microarrays data increases the dimension of the problem to the extent that conventional statistical methods may not be appropriate. In order to deal with the problem of information overflow from the complete sequence of all the genes in a genome using microarrays technique, a combination of hardware together with software solution is needed," Prof. Zee explained.
Although this study involves small number of patients (around 40 to 100), and there is a huge data set with each record consisting of about 20,000 variables on gene expressions. The researcher would then evaluate computer intensive approaches to analyze the data with some statistical models like neural network, classification and regression tree, multi-dimensional mapping, etc. The analysis will be done to locate which gene expressions would have the most effect in causing liver cancer. Further development of methods will also be done in SAS. "We choose SAS because of their years of experience in the life science and the pharmaceutical areas and the sophistication of its technology. SAS has the commitment to continuous progression and investment in R&D on technology that fits our field. I'm glad that SAS technology can support us in the development of statistical and computationally intensive approaches to analyze the genomic data, especially those generated by cDNA micro-array analysis." Prof. Zee commented. "It's important to have the technology that can minimize the research time by not requiring us to do too much customization in the programming. SAS will help to make the process of simulations more efficient in my research under the PC clustering architecture."
In the future, CCTU will continue many research programs, e.g. Bioinformatics projects, clinical trials data management and analysis. "We've been using SAS for more than five years and our unit will explore more opportunities with SAS in the research areas," Prof. Zee concluded.
About Professor Benny Chung-Ying Zee
Professor, Department of Clinical Oncology, The Chinese University of Hong Kong
Dr. Benny Zee joined the Department of Clinical Oncology of the Chinese University of Hong Kong as a Professor and Director of the Comprehensive Cancer Trials Unit in 2001. He has a cross-appointment with the Department of Statistics in the Chinese University of Hong Kong. Recently, he has been appointed as the Director of the Centre for Clinical Trials under the School of Public Health. Professor Zee obtained his Ph.D in Biostatistics from the University of Pittsburgh, USA in 1987. Professor Zee has strong interest in all aspects of multi-center clinical trials, including organizational, statistical and data management aspects. On the methodological side, he has been active in Biostatistics research, especially in the area of quality of life data analysis, bioinformatics, phase II/III trials design, survival analysis, data & safety monitoring and other statistical issues related to clinical trials.
About Comprehensive Cancer Trials Unit (CCTU)
The Comprehensive Cancer Trials Unit (CCTU) was established in early 2001. It is under the Department of Clinical Oncology at the Chinese University of Hong Kong. Its mission is to undertake research that contributes to the reduction of the incidence, morbidity and mortality from cancer. The Unit aims to do this by initiating and supporting different phases of clinical cancer trials, supportive care research and translational studies through grants, donations and contracts with the pharmaceutical industry.
|
.gif){kind=spacer}
