學 術 演 講
學 術 演 講
題 目：Computational Multi-Omics Analysis to Reveal Cancer Potential Causal Factors in Precision Medicine
時 間：民國110年3月22日 (星期一) 上午11：00
Multi-omics data are data from genome, transcriptome, epigenome, proteome, etc. Currently, high throughput technologies including next generation sequencer (NGS), Liquid chromatography-tandem mass spectrometry (LC-MS/MS) etc., enable scientists to measure multi-omics data at once and generate huge amount of data. Multi-omics data provide opportunities for scientists to answer complex questions in biology as well as in the precision medicine. However, huge data size, complex pipeline of statistical/computing analysis, and lack of sufficient domain knowledge reach the limitations of multi-omics approaches in the precision medicine. In this study, genome, transcriptome, proteome data were obtained from cancer and adjacent normal tissues in 91 non-smoking lung adenocarcinomas with early stage. Because total data volume was huge (100 TB per 100 patients), high performance computing servers (Taiwania) from National Center for High-Performance Computing (NCHC) and optimized analytic pipelines were integrated and performed to explore potential causal factors and prognostic patient stratification in non-smoking lung adenocarcinoma. Three major findings 1. endogenous carcinogen, 2. environmental carcinogen were contributed in tumorigenesis, and 3. protein molecular stratification for prognostic risk were demonstrated in this study. The endogenous carcinogen named APOBEC mutational signature was obtained in young female patients with no cancer driver mutation EGFR and patients with APOBEC signature had poor prognosis. By contrast, APOBEC positive patients had better response to cancer immune-therapy. Other risk factors such as nitrosamine and nitro-PAH form environment also played roles in tumorigenesis. Above 2 environmental carcinogens were unique in Taiwan lung cancers compared to the Caucasian patients. Finally, the information from proteome showed the ability to do the molecular stratification for prognostic risk. This is evidence demonstrated that environmental carcinogens played roles in tumorigenesis worldwide and published on the top rank journal Cell. This study is a part of the Taiwan Cancer Moonshot Project which is a member of International Cancer Moonshot Project coordinated by National Cancer Institute US and initiated from current US President Joseph Biden in 2016. Taiwan Cancer Moonshot Project are still working on the muti-omics analysis in breast cancer, pancreatic cancer, stomach cancer, and ovarian cancer, respectively. Computational multi-omics approaches are powerful tools in cancer discovery and it also play important roles for developments of the precision medicine. In the artificial intelligence medicine era, how to use multi-omics information to help patient management or health promotion in the general population are also recognized in the government plan (跨部會健康大數據永續平台).