肺癌在全球與台灣一直是致死率極高的疾病，儘管目前有許多針對肺癌病患所發展的治療策略，但病人存活率仍然不高。而癌轉移則是導致癌症病患死亡的主因。過去研究大多以靜態培養方式來測量細胞侵襲能力，這種培養方式雖然可以大量的培養細胞，但無法貼近真實生理環境，因此，機械仿生動態培養對於肺癌轉移的研究是必要的。本研究將建立機械仿生動態培養環境中細胞侵襲力與基因蛋白體資料庫，並開發微流道晶片做為偵測肺癌轉移相關基因方法，並結合深度學習方法分析TCGA資料庫中肺癌多種基因體資料找出與無復發存活期相關之標的基因。最後，我們將評估這些基因作為預測肺癌病患存活率與復發率之效用。 ;Lung cancer is the most common cause of cancer deaths in Taiwan and worldwide. Although there are several strategies of treatments for lung cancer, the prognosis and 5-year survival rate of patients remain low due to the cancer metastasis. Cancer metastasis is the major cause leading to mortality for cancer patients. Almost cancer metastasis studies were in static cell culture condition. A lot of cells can be harvested in conventional culturing process. Nevertheless, different drawbacks, such as poor cell quality and less physiological-alike biologic characteristics have been noted among the cells cultured by this static condition. Thus, biomimetic dynamic cell culturing process is needed in the field of cancer metastasis research. This project will generate the invasion database of lung cancer cell lines in biomimetic dynamic cell culturing. The database included the transcriptomic profile by RNAseq and proteomic profile by mass. We used microfluidic chip to detect the invasion-associated genes. We also used deep learning-based autoencoder modeling to integrate multiple-omics data and survival data in TCGA lung cancer cohort. Finally, we will validate potential biomarkers in lung cancer patients.
