LncRNA Function Prediction Based On Disease-related SNPs/ceRNA

As it has been found that long noncoding RNAs(lncRNAs)play important roles in a variety of biological processes such as cell proliferation and differentiation,and are associated with many diseases and comorbidities,the functional annotation of lncRNAs in diseases has attracted much attention.However,due to the the large number of false positives tend to generated in predicting lncRNA function by traditional co-expression based analysis methods and the extensive genetic heterogeneity in diseases,the number of lncRNAs whose function is accurately identified by traditional experimental is very limited.Moreover,these problems are more significant when it comes to the functional characteristics of lncRNAs in comorbidities.Therefore,it is crucial to develop a new approach to integrate disease associations to reduce the false discovery rate of lncRNA functional annotations.In this study,disease-related lncRNAs were identified through the integration of single nucleotide polymorphism(SNP)and condition-specific expression,and further combined with cis-regulatory network between lncRNAs and neighboring protein-coding genes as a new strategy(DAnet)to predict the function of lncRNA in disease and reduce the false positives prediction.In addition,there are many diseases with few associated SNPs,such as pregnancy disease of recurrent implantation failure(RIF).By research,only about 30 SNPs were identified as RIF-related,while the relevant lncRNAs could not be identified by DAnet.Lnc RNAs can serve as competing endogenous RNAs(ceRNAs)to adsorb mi RNAs and further regulate their effect on degradation of target mRNAs.Therefore,a competitive endogenous RNA(ceRNA)regulatory network was constructed with differentially expressed lncRNA,mi RNA and mRNA to identify potential regulatory factors and therapeutic drugs in RIF.1.Integrate multiple algorithms to reduce the false positives rate of lncRNA predictionTo identify disease-associated lncRNAs,this study developed a novel strategy DAnet to characterize the function of lncRNAs in diseases by(1)integrating the 193 disease-associated SNPs standardized by the International Classification of Diseases(ICD-11),(2)conditional specific expression of lncRNAs,and(3)the weighted co-expression network of lncRNAs and their neighboring protein-coding genes.After diseaseassociated lncRNAs were identified through disease-related SNPs and conditional specific expression,the optimal conditional specific expression threshold and the chromosomal distance of neighboring genes were screened to identify the more likely disease-related lncRNAs.Furthermore,the gold standard analysis of lncRNA verified by experiments was used to systematically compare the performance of DAnet with the traditional differential expression based approach.The results show that DAnet can characterize the function of lncRNAs in diseases and comorbidities by controlling the false discovery rate and disease heterogeneity.2.Construction of online tool for lncRNA function predictionBased on disease-related SNPs,conditional specific expression,and cis-regulatory networks,this study developed a new strategy(DAnet)to predict the function of lncRNAs in diseases,and built an online tool: FCCLnc for lncRNAs function prediction based on integrated methods for the first time.FCCLnc not only can provides prediction of lncRNA function at optional thresholds,but also can predicts function in simple diseases and comorbidities.In addition,FCCLnc can provide interactive visualization and complete download of lncRNAs centered co-expression networks.Conclusion: In summary,FCCLNC is unique in characterizing the function of lncRNAs in a variety of diseases and comorbidities,and is expected to become an indispensable complement to other available tools.3.Prediction of lncRNA function and disease mechanism by ceRNA networkIn this study,RIF-related sequencing data were collected,and potential regulatory factors and therapeutic drugs of RIF were explored based on ceRNA regulatory network of differentially expressed lncRNA,differentially expressed micronucleotide(micro RNA,mi RNA)and differentially expressed mRNA,and CMap drug analysis.It was found that 10 lncRNAs might as ceRNAs to capture 2 mi RNAs,and then regulate the expression of 30 mRNAs.Amiprilose and dapsone,which have antiinflammatory and immunomodulatory properties,may be candidates for RIF treatment.This study provides a theoretical basis for studying the mechanism of RIF and the application of amiprilose and dapsone in RIF patients from the perspective of lncRNA-mi RNA-mRNA network.In conclusion,this study is the first to integrate multiple methods to characterize the function of lncRNAs in diseases with reduced false discovery rate,and construct an online tool for integrating method for the first time.Further prediction of potential targets and therapeutic drugs of RIF also suggested the important role of lncRNAs in the occurrence of disease.Therefore,the use of bioinformatics methods to predict potential diseaserelated lncRNAs is essential for revealing the pathogenesis of disease,developing new drugs and optimizing personalized medicine.