A novel classification predicts prognosis and drug sensitivity in osteosarcoma based on alterations in gene sets

Abstract

Osteosarcoma is a cancer originating in the bone cells, specifically in the osteoblasts. Previous studies mainly focused on particular molecules but the whole pathway network. We comprehensively analyzed the enrichment score of each signal pathway and identified a novel classification by 20 machine learning algorithms. Furthermore, differences in tumor immune infiltration cells and drug sensitivity were compared in low and high groups. We identified a model consisting of four signaling pathways that predict the prognosis and the immune status of the tumor microenvironment and drug sensitivity in osteosarcoma patients. The novel classification may be used in clinical applications to predict prognosis and drug sensitivity.