Machine learning-driven decoding of maternal immune signatures in repeated pregnancy loss.

Machine learning uncovers maternal T-cell dysregulation and identifies CXCR4 and JUN as key targets in euploid recurrent pregnancy loss.

• Why it matters: Understanding immune mechanisms behind RPL is crucial because immune tolerance failure is a major but poorly characterized factor in pregnancy failure, especially in euploid cases.
• What they did: Single-cell RNA sequencing of decidual tissues from RPL patients and controls was combined with genotype-based origin analysis and advanced machine learning models, including scGPT, to identify immune signatures.
• The result: The study highlights maternal T cells as central to RPL-associated immune dysregulation and proposes CXCR4 and JUN as promising druggable targets for future therapeutic development.