Yang et al. developed a prognostic machine learning model for hepatocellular carcinoma (HCC) by utilizing 21 different programmed cell death gene signatures to predict survival and therapeutics response. Comprised of the dominant subtype of liver cancer globally, HCC contains a multitude of causes and risk factors with a high recurrence and mortality rate. Programmed cell death is used as a marker for tumor proliferation, metastasis, and more with pathway dysregulation leading to complex effects. The researchers integrated signatures such as classical apoptosis, autophagy, cuproptosis, ferroptosis, and disulfidptosis to identify a 10-gene signature and appropriate stratify patients into risk groups. High-risk patients were found to have higher immune evasion of tumors and lower checkpoint gene expression in comparison to low-risk patients. The model accurately predicted therapeutic responses with high-risk patients having a higher sensitivity to different DNA-damaging agents compared to their counterpart. Distinct molecular subtypes were identified, demonstrating different associated pathways. The risk scores were found to show significant correlations with tumor immune microenvironment, immunotherapy response, and drug sensitivity, allowing for insights in future treatments in HHC.