請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/78812
標題: | 可偵測結構變異且具高準確拆分功能之長序列映射器 A Long Read Structure-Variation-Aware Aligner with High-Accurate Sequence Splitting |
作者: | 王健安 Chien-An Wang |
指導教授: | 盧奕璋 |
關鍵字: | 結構變異,序列映射,高通量定序,次世代基因定序,奈米孔道定序, structural variation,sequence mapping,next generation sequencing,high-throughput sequencing,nanopore sequencing, |
出版年 : | 2019 |
學位: | 碩士 |
摘要: | 從2000 年開始,有越來越多生物科技公司投入定序技術的研究並推出新的定序方法,我們稱這些方法為次世代基因定序。而目前被認為最有發展潛力的奈米孔道定序技術,透過偵測蛋白質通道的的電流來進行定序,能夠定序較長的序列片段。長序列片段在分析結構變異與單核苷酸多態性上是更有優勢的。
本論文提出一個長序列映射器SINFA-SS,是改良SINFA 而來的。其演算法分為三個階段,分別是種子序列取樣階段、連結階段與擴展階段,每一階段皆有針對含有結構變異的序列片段經過特殊設計,使其在處理這種事件時有辦法順利拆分與映射。再搭配結構變異呼叫器即能順利找到檢體上含有結構變異的種類與位置。 我們以秀麗隱桿線蟲為參考序列合成含有結構變異的奈米孔道資料。在偵測倒位、缺失、插入和連續重複的能力中不管在準確率較低的R7.3 或是準確率較高的R9.4,SINFA-SS 皆優於Minimap2 與NGMLR。而在與各方面皆表現良好的LAST 比較時,我們準確指出倒位的平均比例較LAST 高出4.31%,準確指出插入的平均比例較LAST高出9.44%,顯示了SINFA-SS 對於分析結構變異的優異表現。 Since year 2000, there were more and more biological technology corporations starting to develop new sequencing methods, usually called next generation sequencing. The nanopore sequencing is considered as the most potential method which can determine the bases by detecting currents flowing through protein pores. This method can sequence the longer reads, which are apt to detect and analyze Single Nucleotide Polymorphisms (SNPs) and Structural Variations (SVs) on samples. In this thesis, we propose a long read structure-variation-aware aligner with high-accurate sequence splitting, SINFA-SS, which is improved from SINFA. The algorithm has three stages: seeding stage, linking stage and extending stage, and each stage is designed to handle the reads with a certain type of the SV, and make reads splitting and mapping. Moreover, we can use SV caller to indicate the type and the positions of the SV on the sample. We use the Caenorhabditis elegans as the reference genome, and simulated the nanopore sequencing reads with SVs. In the ability of detecting inversions, deletions, insertions and tandem duplications, the SINFA-SS is much powerful than Minimap2 and NGMLR whether the type of test data is R7.3 or R9.4. When SINFA-SS is compared with LAST, which is good in all aspects, the average ratio of precisely indicating the inversions by SINFA-SS is 4.31% higher than the one by LAST, and the average ratio of precisely indicating the insertions by SINFA-SS is 9.44% higher than the one by LAST. This shows that SINFA-SS performs well in detecting SVs. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/78812 |
DOI: | 10.6342/NTU201900020 |
全文授權: | 未授權 |
電子全文公開日期: | 2024-02-18 |
顯示於系所單位: | 電子工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-107-1.pdf 目前未授權公開取用 | 3.97 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。