研究主題:利用參考體與配對體個體化健康與疾病預測
撰寫:學生邱千容
2002年發生嚴重急性呼吸道症候群(SARS),而2019年11月中國武漢爆發已知第一例的新型冠狀病毒肺炎疫情,台灣第一例新冠疫情發生於2020年1月21日,大幅改變了民眾的生活型態。
劉子毓老師在2016年曾經在BMC Bioinformatics上發表一篇關於「利用參考體與配對體個體化健康與疾病預測」[1],該篇比較四種病毒(H3N2、H1N1、HRV、RSV)分別在四種狀況下(有基準參考、無基準參考、限制標準預測、差異化預測)的錯誤率以及選擇的基因數量。經由下圖表2中可以看到,無論是哪種病毒,在有基準參考的情況下其錯誤率以及所需要選擇的基因數都較其他三種狀況低。同時,在圖8中可以看到在有參考的情況下,其四種病毒交集出來的共同基因相較沒有參考(圖8左)更為準確、較少。
最終,此篇顯示除了利用參考體在病毒基因判讀上有顯著的進步,該架構未來可以應用在醫療診斷、藥物開發、生物上,使其具有更高的準確度。
In 2002, a severe outbreak of Severe Acute Respiratory Syndrome (SARS) occurred, and in November 2019, the first known case of a novel coronavirus pneumonia epidemic happened in Wuhan, China. The first case of COVID-19 in Taiwan was reported on January 21, 2020, which significantly altered the public’s lifestyle.
Professor Tzu-Yu Liu published a paper in BMC Bioinformatics in 2016, titled “An individualized predictor of health and disease using paired reference and target samples.”[1] This study compared the error rates and the number of selected genes for four viruses (H3N2, H1N1, HRV, RSV) under four scenarios (with reference, without reference, restricted standard prediction, and differentiated prediction). As shown in Table 2 below, for all virus types, the error rates and the number of selected genes were lower when there’s a baseline reference compared to the other three conditions. Additionally, Figure 8 illustrates that the common genes identified among the four viruses were more accurate and fewer when references were used compared to when there were none (left side of Figure 8).
In conclusion, this study indicates that utilizing reference individuals significantly enhances the accuracy of viral gene interpretation, and this framework holds potential for application in medical diagnostics, drug development, and biology in the future, allowing for greater accuracy.