Single-cell and deep learning identify hypoxia-responsive lncRNAs predicting outcomes in colorectal cancer - PubMed
2 hours ago
- #prognostic signature
- #colorectal cancer
- #hypoxia lncRNA
- Hypoxia-responsive lncRNA molecular signature (HRLPMS) predicts outcomes in colorectal cancer (CRC) across multiple data types, including single-cell and spatial transcriptomics.
- HRLPMS is linked to poor survival, chromosomal instability, and specific TME features like fibroblast enrichment, validated by machine and deep learning models.
- High-HRLPMS cells exhibit angiogenesis and ECM remodeling, while low-HRLPMS cells show immune-inflammatory traits, with APP-CD74 interactions potentially driving immunosuppression.
- High-HRLPMS patients may benefit more from treatments like fluorouracil plus bevacizumab and preoperative chemoradiotherapy, highlighting its therapeutic relevance.