iGATTLDA: Integrative graph attention and transformer-based model for predicting lncRNA-Disease associations.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Long non-coding RNAs (lncRNAs) play critical roles in biological processes.
• Dysregulation of lncRNAs is implicated in numerous human diseases.
• Accurate prediction of lncRNA-disease associations is vital for medical advancements.

Purpose of the Study:

• To develop a novel computational method for predicting lncRNA-disease associations.
• To improve the accuracy of identifying links between lncRNAs and human disorders.

Main Methods:

• Constructed a heterogeneous network using lncRNA and disease similarity matrices.
• Employed Graph Attention Network (GAT) to capture local network features.
• Utilized a transformer to model global dependencies in lncRNA-disease interactions.

Main Results:

• The iGATTLDA model achieved high predictive performance.
• Demonstrated an Area Under the ROC Curve (AUC) of 0.95.
• Achieved an Area Under the Precision-Recall Curve (AUPRC) of 0.96.

Conclusions:

• iGATTLDA effectively captures complex lncRNA-disease relationships.
• The model outperforms existing methods in predicting associations.
• This approach enhances disease diagnostics and therapeutic strategies.