Constructing a bladder cancer prognostic model related to exosome using machine learning and identifying THBS1 as a potential target.

Consensus clustering of immune cell proportions classified BCa samples into four distinct subtypes (C1-C4), with differential risk group distributions: C1 (wound healing), C3 (inflammatory), and C4 (lymphocyte-depleted) subtypes were relatively more common in low-risk cases, while the C2 (IFN-γ-dominant) subtype showed a trend toward higher prevalence in high-risk patients, though this association did not reach statistical significance (Fig. 4 E).

also in 53,322 other papers