Title: Function prediction of cancer-related LncRNAs using heterogeneous information network model

Authors: P.V. Sunil Kumar; M. Manju; G. Gopakumar

Addresses: Department of Computer Science and Engineering, National Institute of Technology Calicut, NIT Campus (PO), Kozhikkode, 673601, Kerala, India ' Department of Zoology, KSM Devaswom Board College Sasthamkotta, Kollam, 690521, Kerala, India ' Department of Computer Science and Engineering, National Institute of Technology Calicut, NIT Campus (PO), Kozhikkode, 673601, Kerala, India

Abstract: The aberrant expression of lncRNAs is proven to be one of the prime reasons for cancer progression. Recent studies recommend lncRNAs as potential therapeutic target in cancer. The overexpression of oncogenic lncRNAs causes tumour progression, whereas that of tumour suppressor lncRNAs leads to apoptosis. In this paper, a heterogeneous information network-based Support Vector Machine classifier that can predict lncRNAs into oncogenic or tumour suppressor is proposed. Interactions of lncRNAs with other lncRNAs and proteins along with protein-protein interactions are used to build the network. The model predicted lncRNAs into oncogenic or tumour suppressor with an accuracy of 0.83 and produced an accuracy of 0.80 during an independent validation. A comparison with recently reported studies shows that prediction results fall in line with them.

Keywords: LncRNA; cancer; heterogeneous information network; meta-path; classification; support vector machine; machine learning; oncogenic; tumour suppressor.

DOI: 10.1504/IJDMB.2018.098940

International Journal of Data Mining and Bioinformatics, 2018 Vol.21 No.4, pp.315 - 338

Received: 12 Feb 2018
Accepted: 12 Jan 2019

Published online: 30 Mar 2019 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article