Sztromwasser Paweł, Skrzypczak Damian, Michalak Arkadiusz, Fendler Wojciech
Department of Biostatistics and Translational Medicine, Medical University of Lodz, Łódź, Poland.
Biostatistics Group, Department of Genetics, Wrocław University of Environmental and Life Sciences, Wrocław, Poland.
Front Genet. 2021 Mar 5;12:638960. doi: 10.3389/fgene.2021.638960. eCollection 2021.
Analysis of variants in distant regulatory elements could improve the current 25-50% yield of genetic testing for monogenic diseases. However, the vast size of the regulome, great number of variants, and the difficulty in predicting their phenotypic impact make searching for pathogenic variants in the regulatory genome challenging. New tools for the identification of regulatory variants based on their relevance to the phenotype are needed.
We used tissue-specific regulatory mapped by ENCODE and FANTOM, together with miRNA-gene interactions from miRTarBase and miRWalk, to develop Remus, a web application for the identification of tissue-specific regulatory regions. Remus searches for regulatory features linked to the known disease-associated genes and filters them using activity status in the target tissues relevant for the studied disorder. For user convenience, Remus provides a web interface and facilitates in-browser filtering of variant files suitable for sensitive patient data.
To evaluate our approach, we used a set of 146 regulatory mutations reported causative for 68 distinct monogenic disorders and a manually curated a list of tissues affected by these disorders. In 89.7% of cases, Remus identified the regulator containing the pathogenic mutation. The tissue-specific search limited the number of considered variants by 82.5% as compared to a tissue-agnostic search.
Remus facilitates the identification of regulatory regions potentially associated with a monogenic disease and can supplement classical analysis of coding variations with the aim of improving the diagnostic yield in whole-genome sequencing experiments.
对远距离调控元件中的变异进行分析，可提高目前单基因疾病基因检测25%-50%的阳性率。然而，调控基因组规模巨大、变异数量众多，且预测其表型影响存在困难，使得在调控基因组中寻找致病变异具有挑战性。因此需要基于与表型相关性来鉴定调控变异的新工具。
我们利用由ENCODE和FANTOM绘制的组织特异性调控图谱，以及来自miRTarBase和miRWalk的miRNA-基因相互作用关系，开发了Remus，这是一个用于鉴定组织特异性调控区域的网络应用程序。Remus搜索与已知疾病相关基因相连的调控特征，并根据与所研究疾病相关的靶组织中的活性状态对其进行筛选。为方便用户使用，Remus提供了一个网络界面，并便于在浏览器中对适合敏感患者数据的变异文件进行筛选。
为评估我们的方法，我们使用了一组146个调控突变，这些突变被报告可导致68种不同的单基因疾病，并人工整理了一份受这些疾病影响的组织列表。在89.7%的病例中，Remus鉴定出了包含致病变异的调控因子。与不考虑组织特异性的搜索相比，组织特异性搜索将所考虑的变异数量减少了82.5%。
Remus有助于识别可能与单基因疾病相关的调控区域，并可补充编码变异的经典分析，以提高全基因组测序实验中的诊断阳性率。
