LDAI-ISPS: LncRNA–Disease Associations Inference Based on Integrated Space Projection Scores

Abstract

Long non-coding RNAs (long ncRNAs, lncRNAs) of all kinds have been implicated in a range of cell developmental processes and diseases, while they are not translated into proteins. Inferring diseases associated lncRNAs by computational methods can be helpful to understand the pathogenesis of diseases, but those current computational methods still have not achieved remarkable predictive performance: such as the inaccurate construction of similarity networks and inadequate numbers of known lncRNA–disease associations. In this research, we proposed a lncRNA–disease associations inference based on integrated space projection scores (LDAI-ISPS) composed of the following key steps: changing the Boolean network of known lncRNA–disease associations into the weighted networks via combining all the global information (e.g., disease semantic similarities, lncRNA functional similarities, and known lncRNA–disease associations); obtaining the space projection scores via vector projections of the weighted networks to form the final prediction scores without biases. The leave-one-out cross validation (LOOCV) results showed that, compared with other methods, LDAI-ISPS had a higher accuracy with area-under-the-curve (AUC) value of 0.9154 for inferring diseases, with AUC value of 0.8865 for inferring new lncRNAs (whose associations related to diseases are unknown), with AUC value of 0.7518 for inferring isolated diseases (whose associations related to lncRNAs are unknown). A case study also confirmed the predictive performance of LDAI-ISPS as a helper for traditional biological experiments in inferring the potential LncRNA–disease associations and isolated diseases.

A Novel lncRNA-Disease Association Prediction Model Using Laplacian Regularized Least Squares and Space Projection-Federated Method

“…Thus far, state-of-the-art models for lncRNA-disease association include NCPLDA [33] , IIRWR [25],and LDAI-ISPS [34]. The performance of LRLSSP was assessed by comparing it with three methods in the framework of LOOCV.…”

“…If two diseases share more disease items than others, the similarity between the two diseases would be remarkable. The data were obtained directly from the works of Zhang [34] , Li [33] and YU [25] (Supplementary Materials S10,S11,and S12) and represented by matrix ()…”

“…Finally, a network space projection is used to refine the lncRNAdisease association prediction effect. The area Under receiving operating characteristic (ROC) Curve (AUC) obtained by implementing LOOCV in three datasets is 0.9377 , 0.9383 and 0.9467, respectively,which outperforms those of the three state-of-the-art methods, namely, NCPLDA [33] , LDAI-ISPS [34], and, IIRWR [25],and to demonstrate the effectiveness of LRLSSP. Moreover, LRLSSP was utilized in case studies of melanoma,cervical cancer, ovarian cancer and breast cancer to further validate its effectiveness.…”

“…Chen et al [24] proposed an improved random walk with restart algorithm, Li et al [28]proposed an improved local random walk method for lncRNA-disease association prediction to eliminated the limitations of traditional restart random walk algorithm. Fan et al [29] first used positive pointwise mutual information to construct a large-scale lncRNA-disease heterogeneous network,and then implemented Random Walk with restart algorithm to predict lncRNA disease associations; Hu et al [30] used interaction profile and gene ontology information to build lncRNA-disease network, and then utilized bi-random walks to predict lncRNA-disease associations.HyperGeometric distribution [10], label propagation algorithm [31], KATZ [32], and space projection [33][34][35]have been used to predict lncRNA-disease association, but although their accuracy is low. Xiao et al [36] designed a method based on a simple path with limited length in a heterogeneous network to predict disease-related lncRNAs, but given the extremely simple short-decay function, its prediction performance should be improved with a machine learning method.…”

Two-Stage Inference for LncRNA-Disease Associations Based on Diverse Heterogeneous Information Sources

“…Therefore, in this study, a novel computational method that was composed of a Dual Random Walk (DRW) model in the first stage and a Bipartite Network Space Projection (BNSP) model in the second stage has been explored (namely DRW-BNSP) for inferring lncRNA-disease associations. Our DRW-BNSP can be categorized as one member of network-based method family [13,[15][16][17][40][41][42][43][44][45][46][47][48][49][50][51][52][53]. Existing network-based methods which were similar as ours, such as RWRHLD [54], KATZLDA [55], GrwLDA [56] and IDHI-MIRW [10], used diverse heterogeneous information sources to construct global network models with which to predict potential lncRNA-disease associations.…”