关键词: 长链非编码RNA, 小开放阅读框, 短肽, 表示学习, 深度森林, 预测

Abstract: Abstract: Small open reading frames (sORFs) in long non-coding RNA (lncRNA) can encode short peptides with length no more than 100 amino acids. Aiming at the problem that the features of sORFs in lncRNA are not distinct and the data with high reliability are not enough in short peptide prediction research, a deep forest model based on representation learning is proposed. Firstly, the conventional lncRNA feature extraction method was used to encode the sORFs. Secondly, the autoencoder was used for representation learning to obtain more effective features. Finally, a deep forest model was trained to predict the short peptides encoded by lncRNA. The experimental results show that the accuracy rates of this method can achieve 92.08% on Arabidopsis thaliana dataset, which is higher than that of the traditional machine learning models, deep learning models and combined models, and has better stability. In addition, the prediction accuracy rates of Glycine max and Zea mays dataset can reach 78.16% and 74.92%, which verifies the good generalization ability of the model.

Key words: Keywords: lncRNA, small open reading frames, short peptides, representation learning, Deep Forest, prediction