Higher-Order Dynamic Disentangled Intent Sensing and Bidirectional Joint Updating Framework for NcRNA-Drug Resistance...

Area of Science:

• Pharmacogenomics
• Computational Biology
• Drug Discovery

Background:

• Noncoding RNAs (ncRNAs) significantly influence drug response and disease progression.
• Current prediction methods for ncRNA-drug resistance associations lack specificity and fail to capture complex interactions.
• Existing models often overlook the distinct biological roles and asymmetric relationships between ncRNAs and drugs.

Purpose of the Study:

• To develop an advanced framework, HDBI, for accurate prediction of ncRNA-drug resistance associations.
• To address the limitations of existing methods by incorporating higher-order, dynamic, and disentangled learning.
• To improve the integration of modality-specific and cross-modal features for robust predictions.

Main Methods:

• Proposed HDBI, a higher-order dynamic disentangled framework.
• Integrated multiview hypergraph learning to capture heterogeneous topological patterns.
• Employed disentangled representation modeling and bidirectional cross-modal updating for feature integration.

Main Results:

• HDBI demonstrated superior performance over state-of-the-art methods on benchmark datasets.
• Case studies validated predictions for 5-FU and Docetaxel, with high concordance with existing literature.
• Functional enrichment and molecular docking analyses confirmed the biological relevance of predicted associations.

Conclusions:

• HDBI offers an effective and interpretable approach for identifying ncRNA-mediated drug resistance.
• The framework aids in prioritizing ncRNA targets for drug resistance mechanisms.
• Findings guide downstream mechanistic investigations in pharmacogenomics and precision medicine.