Lactylation-Related Gene LILRB4 Predicts the Prognosis and Immunotherapy of Prostate Cancer Based on Machine Learning

Abstract

Lactylation plays a pivotal role in the metabolic reprogramming, proliferation, migration and immune evasion of tumour cells. However, its specific impact on prostate cancer (PCa) remains poorly understood. This study aimed to investigate the role of lactylation related genes (LRGs) in PCa. LRGs were identified and analysed using data from The Cancer Genome Atlas (TCGA), DKFZ2018, GSE46602 and GSE70768 cohorts. Unsupervised clustering was employed to categorise patients with PCa into two distinct clusters. Prognostic models for PCa were developed using multiple machine learning techniques. LRGs signature was established and validated through training and validation sets. The role of leukocyte immunoglobulin-like receptor B4 (LILRB4) in PCa was examined both in vitro and in vivo. Analysis of LRG expression and prognosis in patients with PCa revealed two distinct clusters with differing survival rates and immune responses. Machine learning models demonstrated the ability to predict survival risks, potentially aiding in the development of personalised treatment strategies. Additionally, LILRB4, a key LRG, promotes PCa progression by modulating the NF-κB and PI3K/AKT pathways, highlighting its potential as a therapeutic target. LRGs exert a pivotal influence on PCa, impacting patient prognosis, immune response and drug sensitivity. The LRGs signature emerges as an essential prognostic tool and a promising therapeutic target for PCa.

Keywords: LILRB4; lactylation; machine learning; prognosis; prostate cancer.