Study On Calculation Method Of Lncrna Functiona Similarity Based On Heterogeneous Network Data

Long non-coding RNAs(long ncRNAs,lncRNAs)are one of great importance noncoding RNA with exceeding 200 nucleotides.Although lncRNAs lack the potential to be translated into proteins directly,their complicated and diversiform functions make them as a window into decoding the mechanisms of human physiological activities.Therefore,exploring the regulation and control functions of lncRNAs is a vital content in the research of non-coding RNAs.With the rapid development of sequencing technique and the continuous progress of bioinformatics,are growing rapidly.However,lncRNAs that have been proved to be functional by experiments are still very limited due to the time and economic costs of biological experiments are expensive.It is urgent to build effective computational models for rapid predicting of unknown lncRNAs functions on a large scale.In the study of bioinformatics,the method that infer the unknown from the known by calculating the similarity between two genes has been widely recognized.Building functional similarity network is a common method to research the correlation between nodes.Therefore,the construction of lncRNAs similarity network is a feasible method of inferring the potential function of lncRNAs.Although there are currently several models that can calculate the similarity between lncRNAs,one common drawback is that they only consider one of many regulatory mechanisms of lncRNAs,which is difficult to universally describe the functional similarity of the lncRNAs.Aiming at this problem,this paper has developed a new method for calculating the functional similarity of lncRNA based on the integration of heterogeneous network data.This model first constructed four single networks like miRNA-based similarity network,disease-based similarity network,GTEx expression profile-based similarity network and NONCODE expression profile-based similarity network.Next,we distinguish positive and negative samples based on whether the lncRNA pairs have common target m RNAs.Then,calculating the AUC of each network as the weight,and using the weighted average of the similarity of four networks as the result of integrated lncRNA function similarity network.Integrated network fully considers the functional similarity of lncRNA in regulating downstream target genes,diseases and the differential expression in organs or tissues,which is more accurate in lncRNA functional similarity describing.After comparing integrated network with single networks and other models,the results show that this integrated network is more effective than any of them.To facilitate researchers to explore the potential functions of lncRNAs,a web server named IHNLnc Sim was constructed for inferring lncRNA functional similarity based on integrating heterogeneous network data.Moreover,this website can visualize the calculation results as a network diagram,which enriches the method of result presentation.In addition,IHNLnc Sim also integrates an lncRNA function enrichment analysis module based on disease associations.It is anticipated that IHNLnc Sim could be an effective bioinformatics tool for the research of lncRNA regulation function studies.