Transcriptomic Profiling of Mouse Mammary Tumors Enables Prognostic and Predictive Biomarker Discovery for Human Breast Cancer.

Elastic Net models trained on baseline gene expression from 26 immunocompetent mouse mammary tumor models predict human breast cancer survival with performance comparable to existing prognostic assays.

• Why it matters: Prognostic and predictive biomarker development is constrained by limited datasets that connect tumor molecular features to treatment response and survival in breast cancer.
• What they did: The study profiled baseline and 7-day on-treatment RNA-seq from mouse models under no treatment, immune checkpoint inhibition, and carboplatin/paclitaxel, then trained Elastic Net predictors and tested them on multiple human datasets.
• The result: Models trained for ICI benefit predicted ICI benefit in human ICI-treated datasets, and 7-day ICI–treated tumor features improved performance, supporting conserved biology for prognosis and ICI response.

AI summary of the abstract · may contain errors · not medical advice · read the paper.