AutoPrimer：一套整合自動化引子設計的Python工具

Abstract

人畜共通病毒指的是那些能透過動物傳播到人類身上的病毒，部分病毒可能只會引發局部的小規模傳染，但也有不少曾引發全球性的大規模疫情。例如：1869年的登革熱疫情、2003年的嚴重急性呼吸道症候群（SARS），以及2019年的新型冠狀病毒（COVID-19）。這些病毒不但危害人體健康，也對整體社會與經濟體系造成了巨大的影響。 病毒的快速檢測對於及早發現疫情、控制病毒傳播，以及制定有效的隔離政策是個重要的環節。在眾多分子檢測技術中，聚合酶連鎖反應（Polymerase Chain Reaction, PCR）仍然被視為診斷病毒感染的一個黃金標準，因為它具備極高的靈敏度與特異性，即便是病毒量極低，也能夠準確檢測出來。而PCR檢測的成敗，大部份取決於引子（primers）設計的好壞，以及其是否具有特異性並能夠準確地結合目標基因序列。 然而，傳統的引子設計流程通常仰賴研究人員手動從基因資料庫中提取目標序列，必要時進行序列比對，並透過不同工具進行設計與特異性分析。這些流程往往操作步驟繁瑣且耗費時間，每一步都依賴人工判斷，也容易導致疏漏或錯誤。當設計目標涵蓋多種病毒序列或是不同基因時，整體效率會大幅下降，難以因應高通量的分析需求。 本研究建立了一套以 Python 編寫的自動化流程，能夠高效率地針對人類病毒基因序列進行引子設計，以因應高通量的分析需求。本系統整合多項工具來自動化執行引子設計，涵蓋四項主要流程：從 NCBI 資料庫提取基因序列，透過 MAFFT 進行多重序列比對，使用 Primer3-py 進行引子設計，最後利用 Primer-BLAST 進行特異性檢驗。此程序會持續進行，直到為每種病毒產生五組通過特異性檢驗的引子，引子的設計相關資訊及其檢驗結果會自動整理成 CSV 檔，以利後續分析。整體而言，與傳統手動方法相比，我們的自動化方法不僅能確保生成出具有特異性的引子對，也大幅提升了批量引子設計流程的效率。

Zoonotic viruses refer to those that originate in animals and have the ability to infect humans. Some of these viruses only cause small, local outbreaks. However, many have led to large and serious global epidemics. Some well-known examples include the dengue virus in 1869, the outbreak of SARS in 2003, and the novel coronavirus (COVID-19) in 2019. These viruses not only threaten human health but also cause major economic and social impacts. Rapid virus detection is an important step in identifying outbreaks early, stopping the spread of infection, and setting effective quarantine policies. Among the many molecular detection methods, Polymerase Chain Reaction (PCR) is still widely seen as the most reliable method for detecting viruses. This is because PCR combines excellent sensitivity with strong specificity, allowing it to accurately detect low levels of viral genetic material. The success of PCR strongly depends on the design of primers that can accurately and specifically target the correct gene sequence. However, traditional primer design is often done manually. Researchers must search for and download gene sequences from databases, perform sequence alignment, design primers using software, and then check their specificity. This process takes a lot of time and effort and also increases the risk of making mistakes. When researchers need to design primers for many viruses or genes, this manual method becomes inefficient and thus unsuitable for large-scale or high-throughput projects. In this study, we present a Python-based automated workflow for high-throughput primer design, focused on human viral genomes. The system automates four main processes: retrieving sequences from NCBI, performing multiple sequence alignment with MAFFT, generating primers using Primer3-py, and checking specificity with Primer-BLAST. This continues until five specific primer pairs are generated for each virus. The final results, including all primer information, are saved in a CSV file for further analysis. Overall, compared to traditional manual methods, our automated approach not only ensures the generation of specific primer pairs but also greatly improves the efficiency of the batch primer design process.