轉譯中心相當榮幸邀請道品股份有限公司共同創辦人 姚文萱博士蒞臨生醫轉譯研究中心演講。

誠摯邀請您的出席參與，

【主辦單位】中央研究院 生醫轉譯研究中心 智慧醫學專題中心
【演講主題】AI 基因體大數據分析改善複雜疾病預測與其他應用
                    Using Big Genomic Data to Improve Complex-Trait Prediction and More with AI.
【演講時間】2022年7月20日 (星期三)
                    上午10:00-11:30 專題演講
                    上午11:30-12:00 交流問答
【演講地點】國家生技研究園區 C212 室
【主講人】姚文萱博士 Wen-hsuang (Adam) Yao, Ph.D.
                 道品股份有限公司 共同創辦人
【引言人】智慧醫學專題中心 林榮信執行長
 

【演講摘要】
     Biomarkers are an emerging field that can potentially guide the diagnosis, prognosis, and treatment course in disease. To identify such biomarkers many artificial intelligence (AI) systems are developed to help finding genetic variations that contribute to common, complex diseases, such as diabetes, cancer, heart disease, asthma, autoimmune disorders, and mental illnesses. However, the immense amount of available data on SNP phenotype associations merely explains a small portion of heritability.
     In this work, we use our recently developed AI-based system, Strata Finder, designed to handle whole genome sequencing (WGS) data so that a vast amount of potentially available information is kept for complex disease analysis. To validate Strata Finder, severe asthma is chosen as our target disease because asthma is a heterogeneous complex disease characterized by reversible airway obstruction, airway hyperresponsiveness, and variable inflammation. Though GWAS have successfully uncovered numerous asthma loci, gaps remain in our understanding of the genetics underlying asthma risk. 
     Severe asthma patients represent a significant portion of healthcare usage (by asthma patients). Here, we present a whole-genome sequencing analysis comparing severe asthma patients to healthy elderly controls. Our results show very good prediction accuracy and several previously reported common asthma related genes are also identified genome-wide. In addition to excellent complex disease risk prediction, our new system has potential to further support medical scientists for exploring patient cohort data to test hypotheses, developing patient stratification schemes, and finding alternative treatment options.
     生物標記能有潛力地運用在疾病的診斷、預後、和治療。為了確認生物標記，許多AI系統發展來尋找複雜疾病基因的變異性，例如：糖尿病、癌症、心臟病、氣喘、自體免疫失調及心理疾病。然而、至今巨量的SNP基因組資料只能解釋小部分的遺傳訊息。
     我們使用自行研發AI系統Strata Finder處理全基因組關聯分析來獲取複雜疾病的成因。嚴重氣喘為我們選擇研究的標靶疾病，因其複雜成因的疾病特性涵蓋可逆的呼吸道阻塞、呼吸道過度反應及多種發炎反應。從全基因組關聯分析成功揭露多種氣喘相關基因組的風險因子。
     嚴重氣喘病人需要重大醫療照護資源。我們使用全基因組關聯分析比較嚴重氣喘病人與健康老人的資料。結果可正確定出造成氣喘相關基因。除了疾病基因風險分析，此系統也可協助醫學科學家探索疾病因子、發展疾病分層及尋找替代治療。
 

【演講提醒】歡迎踴躍參與本演講，為了維護所有與會人員健康安全，請全程配戴口罩。