Node-adaptive graph Transformer with structural encoding for accurate and robust lncRNA-disease association prediction

Abstract

Abstract Background Long noncoding RNAs (lncRNAs) are integral to a plethora of critical cellular biological processes, including the regulation of gene expression, cell differentiation, and the development of tumors and cancers. Predicting the relationships between lncRNAs and diseases can contribute to a better understanding of the pathogenic mechanisms of disease and provide strong support for the development of advanced treatment methods.Results Therefore, we present an innovative node-adaptive Transformer model for predicting unknown associations between lncRNAs and diseases (GNATLDA). First, we utilize the node-adaptive feature smoothing (NAFS) method to learn the local feature information of nodes and encode the structural information of the fusion similarity network of diseases and lncRNAs using Structural Deep Network Embedding (SDNE). Next, the Transformer module, which contains a multi-headed attention layer, is used to learn global feature information about the nodes of the heterogeneous network, which is used to capture potential association information between the network nodes. Finally, we employ a Transformer module with two multi-headed attention layers for learning global-level embedding fusion. Network structure coding is added as the structural inductive bias of the network to compensate for the missing message-passing mechanism in Transformer. Our model accounts for both local-level and global-level node information and exploits the global horizon of the Transformer model, which fuses the structural inductive bias of the network to comprehensively investigate unidentified associations between nodes, significantly increasing the predictive effectiveness of potential interactions between diseases and lncRNAs. We conducted case studies on four diseases; 55 out of 60 interactions between diseases and lncRNAs were confirmed by the literature.Conclusions Our proposed GNATLDA model can serve as a highly efficient computational method for predicting biological information associations.