anamiR:微型核糖核酸與基因表現剖析的整合型分析R套件

Abstract

微型核糖核酸是一群小片段、不轉譯蛋白質的核糖核酸。它們會透過鍵結在目標信使核糖核酸的三端不轉譯區域來抑制後者的轉譯蛋白質，甚至直接降解掉該目標信使核糖核酸。在各種不同的複雜疾病，抑或是不同的病理情況中，基因上的異常、失調是造成發病的因素。因此，在特定的疾病中，如癌症，找出擁有鍵結關係的微型核糖核酸與基因配對是很重要的步驟。然而，要透過生物實驗去驗證這些微型核糖核酸與基因的配對是相當困難的，畢竟配對的數量是如此的龐大，以致我們無法擁有足夠的時間與金錢去一一驗證。近年來，雖然出現了很多關於生物資訊領域所開發的預測演算法，能進行針對微型核糖核酸及基因配對與否的預測，但卻有著各自不同的預測結果，而且彼此間的一致性十分低。因此，我們需要一個有系統性的方法，能夠整合性的同時分析微型核糖核酸以及基因的表現量資料。為了達成這個目的，我們開發了anamiR這個R套件。 anamiR擁有兩個主要的分析流程，能夠結合微型核糖核酸、基因表現量與其樣本對應的表型資訊進行整合性的分析。第一個流程是一般性流程。首先，針對原始的資料進行統計檢定，找出顯著的微型核糖核酸與基因對，再將這些可能存在的配對與一個整合了八個預測演算法，與兩個經生物實驗驗證過的資料集所組合而成的外接資料庫中收集的配對做交集。為了找到可信度高的配對，針對找到的候選目標基因，我們透過建立在涵蓋四種生物途徑資料庫的富集分析，去找到它們可能共同參與的生物功能。針對已經擁有感興趣的基因及或者生物功能的使用者，我們提供第二個流程，基因集合富集分析法流程。這個流程我們將重點擺在感興趣疾病的基因集上。首先，透過基因集合富集分析法，我們先找到了在此疾病中顯著的生物途徑，而一樣透過同上個流程的外接資料庫，我們能夠從已知且顯著的生物途徑中，找到可能參與調控它們微型核糖核酸與基因對。 總結而言，anamiR套件能夠提供整合性的微型核糖核酸與基因表現量資料分析，以及目標基因可能參與到的生物途徑分析。使用者能夠從Bioconductor免費下載anamiR。

MicroRNAs (miRNAs) are small and non-coding RNAs that can regulate gene expression by binding on the 3’UTR of target mRNAs, and also inhibiting mRNAs translating protein, or even promoting mRNAs degradation. In various complex disease and pathological conditions, it is possible to identify dysregulated with causative factors. Therefore, it is an essential approach to explore the interactions between miRNA and gene in certain diseases, such as cancers. However, challenge arises when we are trying to validate the interactions by doing bench experiments. The numbers of miRNA-gene interactions are too large to be validated. Currently, most prediction algorithms only provide their own results and low consistency rates across independent methods have been reported. Consequently, it is necessary to develop a systematic method to perform a comprehensive analysis by using the expression profiles from genes and miRNAs concurrently. To address these issues, a R package named as anamiR was developed. anamiR is able to perform an integrated analysis of mRNA and miRNA with the phenotype information. Two mainly procedures are included. The first one is gGeneral wWorkflow, filtering raw data with statistical test, and comparing the potential miRNA-gene interactions to the embedded databases which contains two validated and eight predicted miRNA-gene databases. To identify potential gene pairs in a specific disease, enrichment analysis based on four pathway databases are applied to obtain putative target genes. For the users who already have interested gene sets or pathways, the other workflow is provided, fFunction dDriven aAnalysis wWorkflow, which allows themus to focus on gene sets in certain diseases , and using embedded databases as well, to identify the miRNA-gene interactions regulating significant term is provided. In summary, the anamiR package provides a comprehensive analysis in expression profiling as well as functional enrichment among miRNAs and their target genes, and is freely available at Bioconductor.