AUTHOR=Yan Cheng , Ding Changsong , Duan Guihua 

TITLE=PMMS: Predicting essential miRNAs based on multi-head self-attention mechanism and sequences

JOURNAL=Frontiers in Medicine

VOLUME=Volume 9 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2022.1015278

DOI=10.3389/fmed.2022.1015278

ISSN=2296-858X

ABSTRACT=Increasing evidences have proved that miRNA plays a significant role in biological progress. Identifying the essential miRNA is helpful for understanding the etiology and mechanisms of various diseases. However, it is time-consuming and expensive to identify essential miRNAs by using traditional biological experiments. It is very urgent to develop computational methods to predict potential essential miRNAs. In this study, we provide a new computational method (called PMMS) to identify essential miRNAs by using multi-head self-attention and sequences. Firstly, PMMS computes the statistic and structure features and extracts the static feature by concatenating them. Secondly, PMMS extracts the deep learning original feature (BiLSTM-based feature) by using BiLSTM (Bi-directional Long Short-Term Memory) and pre-miRNA sequences. In addition, we further obtain the multi-head self-attention feature (MS-based feature) based on BiLSTM-based feature and multi-head self-attention mechanism. By considering the importance of the subsequence of pre-miRNA to the static feature of miRNA, we obtain the deep learning final feature (WA-based feature) based on the weighted attention mechanism. Finally, we concatenate WA-based feature and static feature as input to MLP (Multilayer Perceptron) model to predict essential miRNAs. We conduct 5-fold cross-validation to evaluate the prediction performance of PMMS. The AUC (the areas under ROC curves), F1-score and ACC (accuracy) are used as performance metrics. In according the experiment results, PMMS obtain best prediction performances (AUC: 0.9556, F1-score: 0.9030 and ACC: 0.9097). It also outperforms other compared methods. The experiment results also illustrate that PMMS is an effective method to identify essential miRNA.