Title :
Classification of Real and Pseudo miRNA Precursors Using Local Structure-Sequence Features and Flexible Flexible Neural Tree
Author :
Gaoqiang Yu;Dong Wang;Yuehui Chen
Author_Institution :
Sch. ofInformation Sci. &
fDate :
6/1/2015 12:00:00 AM
Abstract :
MicroRNAs (miRNAs) are a group of short (~22 nt) non-coding RNAs, obtained from the pre-miRNA by nuclease Dicer. MiRNA plays a pivotal regulated role in human disease occurrence, growth and development, cell proliferation and so on. Therefore, miRNA identification has become the primary task in understanding miRNA regulation mechanism. We process pre-miRNA sequence by couplet-syntax and get more detailed description of pre-miRNA sequence that is described by different meaning symbols ("*", "(", "^", "N"). Then, we select the most representative sample characteristics, combine flexible neural tree to predict pre-miRNA. Experimental results show that the average accuracy is 95%.
Keywords :
"Training","Mathematical model","Feature extraction","Probabilistic logic","RNA","Evolution (biology)","Neurons"
Conference_Titel :
Intelligent Computation Technology and Automation (ICICTA), 2015 8th International Conference on
DOI :
10.1109/ICICTA.2015.79