Abstract TU195: Comparison of Large-Scale Proteomics Platforms: The Dallas Heart Study

Introduction: Large-scale proteomic profiling plays an important role in the identification of novel biomarkers. We compared the performance of three proteomics platforms in a multi-ethnic cohort in terms of cross platform agreement and technical variability. Methods: Among community-based participants in the multi-ethnic Dallas Heart Study who participated in the third study phase (2020-2024), we measured proteomics using two affinity-based platforms (SomaLogic 7k; Olink Explore HT) and a mass spectrometry platform (SEER Proteograph) in 186 participants. Eighteen samples were also randomly selected to be run as blind duplicates, which were then used to calculate the coefficient of variation (CV) for each protein. Proteins measured across platforms were matched based on Entrez gene name, and cross-platform Spearman correlations were calculated. Results: Mean age was 60±11 years, 51% were female, 53% reported Black race, 22% reported Asian race, and 7% reported Hispanic ethnicity. The total number of proteins assayed and the number detectable in >50% of samples for each platform were: Olink 5,420 and 2,348 respectively; SomaLogic 7,138 and 7,172 respectively; and SEER 7,796 and 6,831 respectively ( Figure 1A ). Cross-platform correlations were modest, with a median 25 th -75 th percentile range of 0.49 0.20-0.70 for SomaLogic vs Olink; 0.32 0.15-0.51 for SomaLogic vs SEER; and 0.35 0.17-0.49 for SEER vs Olink. Inter-assay CV was lowest for SomaLogic, and for both SEER and Olink, technical variability tended to decrease as the protein detectability increased ( Figure 1B ). Proteins values below the limit of detection in Olink significantly inflated the CV. Conclusions: These three large-scale proteomics platforms exhibit moderate overlap in proteome coverage, and modest agreement in protein quantification for shared proteins assayed. Discovery of novel biomarkers may benefit from verification in multiple proteomics platforms to assess validity of significant hits.