TLSEA 2.0: an updated tool for lncRNA enrichment analysis

Long non-coding RNAs (lncRNAs) regulate gene expression, cellular functions, and disease-related processes through extensive and diverse interactions with various molecules, thereby forming heterogeneous functional regulatory networks. LncRNAs often exert their regulatory effects in a cooperative manner, functional enrichment analysis is considered a powerful approach for systematically elucidating their functions. However, most existing methods focus on the functions of individual lncRNAs, overlooking complex interactions among them, which limits the scope of enrichment analysis. In this paper, we developed TLSEA 2.0 (http://www.lirmed.com/tlsea2/), an updated web-based tool with an expanded repository of functional lncRNA sets, for enhancing enrichment analysis capabilities. Compared with the earlier version of TLSEA, TLSEA 2.0 has tripled its reference lncRNA sets and expanded enrichment analysis categories from one to five: Disease, Drug, RNA-RNA Interaction, RNA-Protein Interaction, and Cancer Phenotype. TLSEA 2.0 introduces an Expansion option, which extends the user-submitted lncRNA list by incorporating additional lncRNAs that exhibit strong association with the input set based on similarity networks, thereby expanding the user-submitted lncRNA list and enabling the discovery of more potential associations. Additional functional similarity networks have been incorporated, including lncRNA-protein interactions and co-expression data. Furthermore, TLSEA 2.0 employs combining Graph Attention Networks (GAT) and Graph Convolutional Networks (GCN) to replace traditional graph representation learning during the process of extracting lncRNA features. These enhancements make TLSEA 2.0 a more comprehensive and robust online platform for functional enrichment analysis, facilitating deeper insights into the complex biological regulatory functions of lncRNAs.