Machine learning identifies a DNA repair-related risk model for bladder cancer and functionally characterizes ZWINT...

Area of Science:

• Oncology
• Genetics
• Bioinformatics

Background:

• Bladder cancer (BCa) exhibits significant molecular and clinical heterogeneity, complicating prognosis and treatment decisions.
• Accurate prognostic assessment and treatment stratification are hindered by this heterogeneity.

Purpose of the Study:

• To develop a machine learning-based prognostic model utilizing DNA repair-related genes for bladder cancer.
• To investigate the functional role of specific genes within this model in BCa progression.

Main Methods:

• A DNA repair-related risk score (DRRS) was constructed using 101 machine learning algorithms on BCa transcriptomic data.
• Prognostic value, immune relevance, and therapeutic implications of DRRS were assessed.
• Functional characterization of ZWINT (ZW10 interacting kinetochore protein) was performed via in vitro experiments.

Main Results:

• DRRS successfully stratified BCa patients into high- and low-risk groups with differing survival, immune microenvironment, and treatment response profiles.
• ZWINT knockdown inhibited cell proliferation, migration, and invasion, induced apoptosis, and caused G1 cell cycle arrest.
• ZWINT knockdown led to increased p65 phosphorylation, suggesting NF-κB pathway activation.

Conclusions:

• A machine learning-derived DNA repair signature (DRRS) for BCa was established.
• ZWINT was functionally validated as a potential oncogenic driver in BCa.
• DRRS and ZWINT represent promising biomarkers and therapeutic targets for bladder cancer treatment stratification.