Using proteomics for stratification and risk prediction in patients with solid tumors

Proteomics, the study of proteins and their functions, has greatly evolved due to advances in analytical chemistry and computational biology. Unlike genomics or transcriptomics, proteomics captures the dynamic and diverse nature of proteins, which play crucial roles in cellular processes. This is exemplified in cancer, where genomic and transcriptomic information often falls short in reflecting actual protein expression and interactions. Liquid chromatography–mass spectrometry (LC-MS) is pivotal in proteomic data generation, enabling high-throughput analysis of protein samples. The MS-based workflow involves protein digestion, chromatographic separation, ionization, and fragmentation, leading to peptide identification and quantification. Computational biostatistics, particularly using tools in R (R Foundation for Statistical Computing, Vienna, Austria; www.R-project.org), aid in data analysis, revealing protein expression patterns and correlations with clinical variables. Proteomic studies can be explorative, aiming to characterize entire proteomes, or targeted, focusing on specific proteins of interest. The integration of proteomics with genomics addresses database limitations and enhances peptide identification. Case studies in intrahepatic cholangiocarcinoma, glioblastoma multiforme, and pancreatic ductal adenocarcinoma highlight proteomics’ clinical applications, from subtyping cancers to identifying diagnostic markers. Moreover, proteomic data augment molecular tumor boards by providing deeper insights into pathway activities and genomic mutations, supporting personalized treatment decisions. Overall, proteomics contributes significantly to advancing our understanding of cellular biology and improving clinical care.