Machine learning on transcription factor expression profiles for precision breast cancer therapy

Area of Science:

• Oncology
• Bioinformatics
• Machine Learning

Background:

• Breast cancer is a heterogeneous disease with increasing global prevalence.
• Accurate prognostic evaluation is crucial for personalized therapy and patient survival.
• Machine learning is redefining cancer prognosis and prediction.

Purpose of the Study:

• To develop and validate a novel Machine Learning-Derived Transcription Factor Signature (MDTS) for breast cancer prognosis.
• To assess the predictive power of MDTS against existing signatures.
• To identify therapeutic strategies based on MDTS scores.

Main Methods:

• A ten-fold cross-validation method was used to construct the MDTS across 108 algorithmic combinations.
• The optimal model was selected based on the highest average C-index across ten cohorts.
• Single-cell and multi-omics data were integrated to assess MDTS robustness.

Main Results:

• The MDTS demonstrated superior predictive power, outperforming 103 existing signatures.
• MDTS accurately predicted breast cancer outcomes across 10 independent cohorts.
• Patients with low MDTS scores may benefit from immunotherapy; high MDTS scores suggest PAC-1 as a targeted chemotherapy agent.

Conclusions:

• The developed MDTS offers a robust tool for breast cancer outcome prediction.
• MDTS facilitates personalized treatment strategies, including immunotherapy and chemotherapy.
• These findings pave the way for advanced MDTS-based breast cancer therapy customization.