針對基因組裝中欠缺的間隙區域以標記性長片段做填補

Abstract

次世代定序（Next Generation Sequencing）因為定序讀長與組裝技術的限制，導致基因體初稿中序列之內或之間產生間隙序列（Gaps），亦即一段未知、尚未被確認的序列。透過解決間隙序列，將可以在組裝中找到更多的基因，不僅可以提高組裝中的基因含量，更能夠達到更完整的基因體組裝。本論文使用 10x Genomics公司所開發之技術，將 Barcode 標記結合短序列定序以獲取長片段訊息（Linked Reads）；嘗試利用其資料特性，因應以往棘手的重複序列（Repeats）問題，透過解析基因體組裝中的間隙序列，進行填補與修復。實驗結果顯示，針對日本鰻組裝中 11.8%的間隙區域有效的下降到 4.5%，並且大幅提高了組裝中的基因含量。

De novo assembly of short reads that accompanies with scaffolding algorithm often produces lots of gaps due to a lack of information to determine the sequences between contigs which leads to an incomplete draft of genomes. Here, we present a protocol for filling the gaps in genome that utilizes linked-read information contained in barcoded short reads. We validate our protocol with Japanese eel genome and show how the gaps can be filled using 10x Chromium linked reads with our local assembly methods. We expect this gap-filling protocol can be utilized for reaching more complete and high-quality scaffolds in genome assembly drafts.