Sadeghi B, Ahmadi H, Azimzadeh-Jamalkandi S, Nassiri M R, Masoudi-Nejad A
Laboratory of Molecular Genetics and Bioinformatics (LMGB), Department of Animal Science, College of Agriculture Science and Natural Resource, Gonbad Kavous University, Gonbad Kavous, Iran; Laboratory of System Biology and Bioinformatics (LBB), Institute of Biochemistry and Biophysics, University of Tehran, Tehran, Iran.
Anim Genet. 2014 Aug;45(4):479-84. doi: 10.1111/age.12170. Epub 2014 May 17.
MicroRNAs (miRNAs) are small non-coding RNAs that modulate gene expression transcriptionally (transcriptional activation or inactivation) and/or post-transcriptionally (translation inhibition or degradation of their target mRNAs). This phenomenon has significant roles in growth and developmental processes in plants and animals. Bos taurus is one of the most important livestock animals, having great importance in food and economical sciences and industries. However, limited information is available on Bos taurus constituent miRNAs because its whole genome assembly has been only recently published. Therefore, computational methods have been essential tools in miRNA gene prediction and discovery. Among these, machine-learning-based approaches are used to characterize genome scale pre-miRNAs from expressed sequence tags (ESTs). In this study, a support vector machine model was used to classify 33 structural and thermodynamic features of pre-miRNA genes. Public bovine EST data were obtained from different tissues in various developmental stages. A new algorithm, called BosFinder, was developed to identify and annotate the whole genome's derived pre-miRNAs. We found 18 776 highly potential pre-miRNA sequences. This is the first genome survey report of Bos taurus based on a machine-learning method for pre-miRNA gene finding. The bosfinder program is freely available at http://lbb.ut.ac.ir/Download/LBBsoft/BosFinder/.
微小RNA（miRNA）是一类小的非编码RNA，可在转录水平（转录激活或失活）和/或转录后水平（抑制靶mRNA的翻译或使其降解）调控基因表达。这种现象在植物和动物的生长发育过程中发挥着重要作用。牛是最重要的家畜之一，在食品和经济科学及产业中具有重要意义。然而，由于牛的全基因组组装最近才公布，关于其组成性miRNA的信息有限。因此，计算方法一直是miRNA基因预测和发现的重要工具。其中，基于机器学习的方法被用于从表达序列标签（EST）中鉴定基因组规模的前体miRNA。在本研究中，使用支持向量机模型对前体miRNA基因的33个结构和热力学特征进行分类。从不同发育阶段的不同组织中获取了公开的牛EST数据。开发了一种名为BosFinder的新算法，用于识别和注释全基因组衍生的前体miRNA。我们发现了18776个具有高度潜力的前体miRNA序列。这是基于机器学习方法进行前体miRNA基因发现的牛的首个全基因组调查报道。BosFinder程序可从http://lbb.ut.ac.ir/Download/LBBsoft/BosFinder/免费获取。
