A machine learning-driven framework integrating cell death and senescence signatures for multi-target drug design and...

Area of Science:

• Oncology
• Genomics
• Immunology

Background:

• Ovarian cancer (OC) presents significant therapeutic challenges due to molecular heterogeneity and adaptive resistance.
• Non-apoptotic cell death and senescence pathways are implicated in tumor evolution and immunosuppression but are underexplored in predictive models.
• Integrating these pathways is crucial for multi-target drug design and immunotherapy optimization in OC.

Purpose of the Study:

• To identify key genes governing cell death and senescence (CDS) in ovarian cancer using machine learning.
• To develop and validate a predictive signature (Cell Death and Senescence Learning Signature - CDSLS) for OC patient outcomes and immune microenvironment.
• To functionally characterize the role of the hub gene RB1 in senescence, DNA damage, and T-cell activation, and explore AI-driven therapeutic targeting.

Main Methods:

• Machine learning algorithms were employed to identify genes associated with cell death and senescence.
• Multi-omics analyses, including single-cell RNA sequencing, were performed to map the tumor microenvironment.
• In vitro and in vivo experiments validated the function of RB1, and AI tools predicted small molecule compounds targeting RB1.

Main Results:

• The Cell Death and Senescence Learning Signature (CDSLS) was validated across multiple OC cohorts (n=1858) and immunotherapy datasets, correlating with poor survival and immunosuppression.
• RB1 knockdown promoted proliferation and suppressed senescence, while RB1 overexpression induced senescence, amplified DNA damage, and enhanced CD8+ T cell activation.
• RB1-overexpressing tumors exhibited restrained growth and increased immune infiltration, with AI predicting targeted small molecule compounds for RB1.

Conclusions:

• The developed CDSLS framework effectively predicts ovarian cancer patient survival and immune status.
• RB1 plays a critical role in regulating senescence, DNA damage response, and anti-tumor immunity, representing a potential therapeutic target.
• This study highlights the utility of AI in identifying conserved therapeutic targets, predicting drug responses, and advancing ovarian cancer treatment strategies.