GeneFriends: An RNA-seq co-expression tool for functional annotation and candidate gene prioritization

在过去的十年中，部分由于人类基因组的测序，遗传学领域的研究迅速扩大，导致与多种人类疾病和特征相关的许多基因的鉴定。尽管科学对大多数人类疾病有了更广泛的了解，但仍有许多未解之谜，大多数人类常见疾病和大多数人类特征的遗传基础只得到部分理解。该项目旨在构建一个新的在线工具，以帮助研究人员推断未知基因的功能，并确定在疾病和生物过程中发挥作用的候选新角色。该工具将基于从最先进的RNA测序技术（RNA-seq）获得的数据。这项技术使研究人员能够以比以前更高的准确度测量给定样本中基因的活性，并且能够测量不编码蛋白质的基因。越来越多的研究人员目前使用RNA-seq，但许多与疾病和过程相关的基因具有未知的功能，这阻碍了对相关机制的理解。为了解决这个问题，我们提出的工具，名为GeneFriends，采用了一种关联内疚的方法来分析基因。这种方法可以识别哪些基因倾向于在多个样本中同时被激活。基于同时被激活的基因在功能上是相关的这一想法，有可能根据与它们共同激活的已知基因的功能来预测以前未研究的基因的功能。为了确定基因是如何被共同激活的，将结合数千个先前使用RNA-seq测量的基因水平的公开样本。这样就可以生成一个图谱，描述哪些基因通常倾向于一起起作用。基因之友将允许科学家和临床医生推断未知基因的功能。此外，GeneFriends还可用于将新基因与疾病和过程相关联，因为它们与先前确定为对该疾病或过程重要的其他基因同时活跃。该项目中创建的相应工具将使研究人员能够将新因素与疾病联系起来，确定候选药物靶点，从而可能导致新型诊断和治疗。GeneFriends最初将为人类和小鼠创建，并将在网上免费提供给研究人员，临床医生和商业公司使用。

Over the past decade, and in part thanks to the sequencing of the human genome, research in the field of genetics has rapidly expanded leading to the identification of many genes associated with multiple human diseases and traits. Even though science has gained a broader understanding of most human diseases, many unanswered questions remain and the genetic basis of the majority of human common diseases and most human traits is only partly understood. This project aims to construct a new online tool to help researchers infer the functions of unknown genes as well as identify candidate new players for roles in diseases and in biological processes. This tool will be based on data obtained from the state-of-the-art RNA sequencing technology (RNA-seq). This technology allows researchers to measure the activity of genes in a given sample with greater accuracy than previous methods and is capable of measuring genes that to not encode proteins. A growing number of researchers currently employ RNA-seq, yet many genes associated with diseases and processes have unknown functions, which impedes understanding the mechanisms involved. To address this issue, our proposed tool, entitled GeneFriends, employs a guilt-by-association methodology to the analysis of genes. This approach identifies which genes tend to be activated simultaneously across multiple samples. Based on the idea that genes that are activated at the same time are functionally related, it is possible to predict the functions of previously unstudied genes based on the functions of the known genes they are co-activated with. To determine how genes tend to be co-activated, thousands of publicly available samples of previously measured gene levels using RNA-seq will be combined. This way a map will be generated that describes which genes generally tend to act together. GeneFriends will allow scientists and clinicians to infer the function of unknown genes. Additionally, GeneFriends can be used to associate new genes with diseases and processes based on the fact that they were active at the same time as other genes previously identified as important for that disease or process. The corresponding tool created in this project will then allow researchers to relate new factors to diseases, identify candidate drug targets, potentially leading to new types of diagnosis and treatments. GeneFriends will be initially created for humans and mice and will be made freely available online for researchers, clinicians and commercial companies to use.