TLSEA

TLSEA performs enrichment analysis of long non-coding RNAs (lncRNAs) by integrating heterogeneous data and network-based methods to identify biological pathways and functional categories associated with genes encoding lncRNAs.

Key Features:

• Multi-Source Heterogeneous Information Fusion: Integrates heterogeneous information from multiple sources to construct a comprehensive lncRNA-lncRNA association network.
• Graph Representation Learning: Applies graph representation learning to extract low-dimensional vectors of lncRNAs within two functional annotation networks.
• Novel Network Construction: Merges various lncRNA-related similarity networks to construct an integrated association network.
• Random Walk with Restart Method: Expands user-submitted lncRNA sets using the random walk with restart algorithm on the lncRNA-lncRNA association network.

Scientific Applications:

• Complex disease mechanism research: Supports investigation of molecular mechanisms in complex human diseases by revealing lncRNA-associated pathways and functional categories.
• Breast cancer case study: Demonstrated improved detection of disease-related pathways and functional categories in a breast cancer case study compared to conventional tools.

Methodology:

Construction of an integrated lncRNA-lncRNA association network from heterogeneous data sources; application of graph representation learning to derive low-dimensional lncRNA vectors within two functional annotation networks; and expansion of user-submitted lncRNA sets via the random walk with restart method on the association network.