Predicting miRNA-disease interaction based on recommend method
| Published date | 06 September 2019 |
| Pages | 35-40 |
| DOI | https://doi.org/10.1108/IDD-04-2019-0026 |
| Date | 06 September 2019 |
| Author | Qingfeng Chen,Zhe Zhao,Wei Lan,Ruchang Zhang,Jiahai Liang |
| Subject Matter | Library & information science,Library & information services,Lending,Document delivery,Collection building & management,Stock revision,Consortia |
Predicting miRNA-disease interaction based on
recommend method
Qingfeng Chen
School of Computer, Electronics and Information, Guangxi University, Nanning, China
Zhe Zhao
Guangxi University, Nanning, China
Wei Lan
School of Computer, Electronics and Information, Guangxi University, Nanning, China
Ruchang Zhang
Guangxi University, Nanning, China, and
Jiahai Liang
School of Electronic and Information Engineering, Beibu Gulf University, Qinzhou, China
Abstract
Purpose –MicroRNAs (miRNAs) have been proved to be a significant type of non-coding RNAs related to various human diseases. This paper aims
to identify the potential miRNA–disease interactions.
Design/methodology/approach –A computational framework, MDIRM is presented to predict miRNAs-disease interactions. Unlike traditi onal
approaches, the miRNA function similarity is calculated by miRNA–disease interactions. The k-mean method is further use d to cluster miRNA
similarity network. For miRNAs in the same cluster, their similarities are enhanced, as the miRNAs from the same cluster may be reliable. Further,
the potential miRNA–disease association is predicted by using recommend method.
Findings –To evaluate the performance of our model, the fivefold cross validation is implemented to compare with two state-of-the-art methods.
The experimental results indicate that MDIRM achieves an AUC of 0.926, which outperforms other methods.
Originality/value –This paper proposes a novel computational method for miRNA–disease interaction prediction based on recommend method.
Identifying the relationship between miRNAs and diseases not only helps us better understand the disease occurrence and mechanism through the
perspective of miRNA but also promotes disease diagnosis and treatment.
Keywords Cluster, K-mean, Cross-validation, miRNA functional similarity, MiRNA–disease association, Recommend method
Paper type Research paper
1. Introduction
MicroRNA (miRNA) is a class of short (approximately 22 nt),
endogenous and single-stranded non-coding RNAs, which are
normally regarded as the gene expressionregulator by targeting
mRNA to control mRNA’s expression in the post-
transcriptional process (Ambros et al., 2004 and Chen et al.,
2016a). Accumulating experimental evidences have
demonstrated that miRNAs play a significant role in various
plants’, animals’andviruses’biological processes (Pfeffer et al.,
2006 and Lan et al.,2015), such as cell proliferation
(Brennecke et al.,2003), development (Baltimore et al.,2008),
differentiation (Martin et al., 2016), apoptosis (Cheng et al.,
2005), metabolism (Bing et al., 2012) and aging (Chen et al.,
2010). The formulation of mature miRNAs consists of three
steps: first, the primary miRNA (pri-miRNA) is clipped by
Drosha nucleotide enzyme and processed into precursor
miRNA (pre-miRNA) with approximately 60nt in length.
Then, the generated pre-miRNA is transported to the
cytoplasm and processed into mature miRNA by RNase III
enzyme with about 22nt in length. Finally, miRNAs are
selectively loaded into an RNA-induced silencing complex to
regulate gene expression (Lander et al.,2001;Lan et al.,2017;
Meister et al., 2004;Chen et al., 2016b).
It has been proved that the dysregulation of miRNA is
associated with large amount of human diseases (Alvarezgarcia
et al., 2005; Lan et al., 2014;Lu et al., 2008). For example,
miR-122 has been proved as the most abundant liver-specific
miRNA in gene networks and pathways of liver diseases
Thecurrentissueandfulltextarchiveofthisjournalisavailableon
Emerald Insight at: https://www.emerald.com/insight/2398-6247.htm
Information Discovery and Delivery
48/1 (2020) 35–40
© Emerald Publishing Limited [ISSN 2398-6247]
[DOI 10.1108/IDD-04-2019-0026]
This paper was partially supported by the National Natural Science
Foundation of China projects 61702122 and 61751314, the Natural
Science Foundation of Guangxi projects 2017GXNSFDA198033 and
2018GXNSFBA281193, the key R&D plan of Guangxi AB17195055, the
Scientific Research Fund of Hunan Provincial Education Department
18B469 and the Doctoral research fund of Guangxi University
XBZ180476.
Received 1 April 2019
Revised 9 May 2019
Accepted 15 May 2019
35
To continue reading
Request your trial