Singh Sumi, Benton Ryan G, Singh Anurag, Singh Anshuman
School of Computer Science and Mathematics, University of Central Missouri, Warrensburg, MO, 64093, USA.
Department of Computer Science, University of South Alabama School of Computing, Shelby Hall, Suite 2101, 150 Jaguar Drive, Mobile, AL, 36688, USA.
Methods Mol Biol. 2017;1617:211-224. doi: 10.1007/978-1-4939-7046-9_16.
In recent years, the role of miRNAs in post-transcriptional gene regulation has provided new insights into the understanding of several types of cancers and neurological disorders. Although miRNA research has gathered great momentum since its discovery, traditional biological methods for finding miRNA genes and targets continue to remain a huge challenge due to the laborious tasks and extensive time involved. Fortunately, advances in computational methods have yielded considerable improvements in miRNA studies. This literature review briefly discusses recent machine learning-based techniques applied in the discovery of miRNAs, prediction of miRNA targets, and inference of miRNA functions. We also discuss the limitations of how these approaches have been elucidated in previous studies.
近年来,微小RNA(miRNA)在转录后基因调控中的作用为理解多种癌症和神经疾病提供了新的视角。尽管自miRNA被发现以来,相关研究发展迅速,但由于任务艰巨且耗时较长,寻找miRNA基因和靶标的传统生物学方法仍然面临巨大挑战。幸运的是,计算方法的进步在miRNA研究中取得了显著进展。这篇文献综述简要讨论了最近基于机器学习的技术在miRNA发现、miRNA靶标预测和miRNA功能推断中的应用。我们还讨论了这些方法在以往研究中存在的局限性。