Construction of endothelial cell signatures for predicting the diagnosis, prognosis and immunotherapy response of bladder cancer via machine learning.

We subtyped bladder cancer (BC) patients based on the expression patterns of endothelial cell (EC) -related genes and constructed a diagnostic signature and an endothelial cell prognostic index (ECPI), which are useful for diagnosing BC patients, predicting the prognosis of BC and evaluating drug sensitivity. Differentially expressed genes in ECs were obtained from the Tumour Immune Single-Cell Hub database. Subsequently, a diagnostic signature, a tumour subtyping system and an ECPI were constructed using data from The Cancer Genome Atlas and Gene Expression Omnibus. Associations between the ECPI and the tumour microenvironment, drug sensitivity and biofunctions were assessed. The hub genes in the ECPI were identified as drug candidates by molecular docking. Subtype identification indicated that high EC levels were associated with a worse prognosis and immunosuppressive effect. The diagnostic signature and ECPI were used to effectively diagnose BC and accurately assess the prognosis of BC and drug sensitivity among patients. Three hub genes in the ECPI were extracted, and the three genes had the closest affinity for doxorubicin and curcumin. There was a close relationship between EC and BC. EC-related genes can help clinicians diagnose BC, predict the prognosis of BC and select effective drugs.