Meta-análise computacional de dados proteômicos de tecidos humanos para identificação de antígenos de câncer-testículo

Abstract

The proteins encoded by the genome constitute the proteome, which regulates various biological processes within an organism. However, the proteome is not limited to the sum of genome-encoded products, it also encompasses protein variations resulting from post-transcriptional and post-translational events. The study of the proteome has been termed proteomics, a promising approach that provides insights into protein levels in various biological and clinical models. In contrast to genomics, the availability of proteomic data remains limited, underscoring the need for consolidated practices for data storage in public repositories. Moreover, proteomic studies tend to be inherently more complex than their genomic. Although this challenge, the field has experienced substantial technological advancements in sensitivity and sequencing capacity. As a result, the volume of data generated by proteomics laboratories has increased significantly, enabling the integration of these datasets through meta-analysis techniques, an efficient strategy for combining data from different studies. Biomarkers are molecular markers found in clinical samples which may aid disease diagnosis or prognosis, including cancer. High-throughput techniques allow prospecting for such signature molecules by comparing gene expression between normal and tumor cells. Cancer-testis antigens (CTAs) are promising candidates for cancer biomarkers due to their limited expression to the testis in normal conditions versus their aberrant expression in various tumors. CTAs are routinely identified by transcriptomics, which limits biomarker characterization, as predictions based solely on the transcriptomic level do not correlate with protein abundance. Thus, it becomes essential to complement transcriptomic data with proteomic analyses in order to avoid false-positive predictions when relying solely on a single omics strategy. In this study, we ran a computational meta-analysis on several proteomic datasets from tumor and healthy human tissues, aiming to validate, at the protein level, the CTAs previously identified by transcriptomic studies, to investigate their abundance in the proteome layer, and evaluate their potential as biomarker candidates. The combined datasets present the expression patterns of 17,200 unique proteins, including 241 known CTAs previously described at the transcriptomic level. Those were further ranked as significantly enriched in tumor tissues (23 proteins), exclusive to tumor tissues (26 proteins) or abundant in healthy tissues (8 proteins). Our study illustrates the possibilities for tumor