SenolyticSynergy: An Attention-Based Network for Discovering Novel Senolytic Combinations via Human Aging Genomics

Area of Science:

• Gerontology and pharmacology
• Computational biology and bioinformatics
• Drug discovery and development

Background:

• Senolytics, drugs targeting aging, have shown promise since 2015.
• Mechanism-based strategies, not random screening, led to early discoveries like dasatinib (D) and quercetin (Q).
• Predicting effective senolytic combinations requires extensive experimental data, limiting current research.

Purpose of the Study:

• To develop a computational framework for predicting synergistic senolytic drug combinations.
• To identify novel, high-confidence senolytic drug combinations for aging interventions.
• To overcome limitations of experimental data requirements in senolytic drug discovery.

Main Methods:

• Differential gene expression analysis and machine learning techniques were employed.
• A multimodal attention mechanism model was retrained on 1624 core aging-related genes, creating the SenolyticSynergy framework.
• 63 senolytic compounds were used to generate 1953 potential drug combinations for analysis.

Main Results:

• SenolyticSynergy identified 190 high-confidence drug combinations with predicted synergistic scores.
• Ten combinations showed exceptionally high synergistic scores (>8).
• The combination of temsirolimus and nitazoxanide ranked highest and is a promising candidate, validated by literature and molecular modeling.

Conclusions:

• The SenolyticSynergy framework enables large-scale prediction of anti-aging drug combinations.
• This approach accelerates research into novel senolytic therapies.
• Temsirolimus and nitazoxanide represent a potentially significant breakthrough in senolytic drug development.