“…In this study, we used six machine learning algorithms to construct a model of prognosis-related DEGs between subtypes, identified 30 genes involved in each machine learning model, and captured six of them to develop a risk evaluator. Each gene in the risk evaluator, PLXNA1 (Ho, 1988), MARCKSL1 (Egeland et al, 2019), IQGAP3 (Leone et al, 2021), PFN2 (Cui et al, 2016), PON1 (Bobin-Dubigeon et al, 2012, and TAK (Li et al, 2022), has been reported to be associated with the prognosis or progression of cancer. Here, the risk evaluator developed using these six genes was associated with potential factors affecting the prognosis of HCC, such as clinicopathological features, somatic mutations, tumor microenvironment indicators, and signaling pathways.…”