J Proteomics. 2026 Sep 19. pii: S1874-3919(26)00138-7. [Epub ahead of print]
105735
Jean-Marc Monneuse,
Hayat Hage,
Célie Da Silva,
Mallory Poncelet,
Benoit Levast,
Karen Louis,
Sangeetha Covindarassou,
Thomas Ribette,
Joséphine Abi Ghanem,
Magali Sarafian,
Adrien Saliou.
Plasma proteomics holds promise for biomarker discovery, yet its characterization remains limited by the complex dynamic range of protein concentrations. We benchmarked seven proteomics workflows including five MS-based strategies (NEAT, PCA-N, TOP14, ENRICH-iST, SEER) on the Orbitrap Astral and Exploris 480, and Olink Target Inflammation and Reveal on 80 lithium heparin plasma samples from two clinical cohorts. Workflows were evaluated for proteome depth, reproducibility, HPPP catalogue coverage and recovery of aging-associated biological signatures. The Astral outperformed the Exploris 480 three-fold in protein identifications at four-fold higher throughput. Proteome coverage ranged from 1159 (NEAT) to 6003 protein groups (SEER), the latter providing over 70% HPPP coverage. TOP14 and ENRICH-iST achieved intermediate depth with high reproducibility. ENRICH-iST further demonstrated applicability to lithium heparin plasma via a simple protocol adaptation. Both Olink assays quantified their complete target panels, capturing low-abundance markers partially inaccessible to MS. Despite limited numerical overlap between MS and Olink datasets, fold-change directionality was broadly concordant. SEER and Olink Reveal uniquely recovered aging-related pathway signatures, including telomere maintenance and immune modulation. No single workflow provides exhaustive plasma proteome coverage. Integration of MS and affinity-based assays should be considered in plasma biomarker research for comprehensive coverage of the plasma proteome. SIGNIFICANCE: Plasma proteomics is central to biomarker discovery and translational research, yet practical adoption of next-generation workflows is hampered by the absence of head-to-head data comparing both mass-spectrometry and affinity platforms on the same clinical samples. Plasma proteomics is central to biomarker discovery and translational research, owing to the accessibility of plasma and its capacity to reflect physiological and pathological processes across tissues. Despite major technological advances in both mass spectrometry-based and affinity-based proteomics, the practical adoption of these innovations in clinical and population-scale studies remains constrained by methodological heterogeneity and the absence of comprehensive comparative frameworks. As a result, study design and platform selection are often driven by technical availability rather than by an informed assessment of analytical trade-offs and biological objectives. To address this gap, we benchmarked seven workflows including five MS-based strategies (NEAT, TOP14 depletion, ENRICH-iST, PCA-N, and Seer Proteograph XT) and two affinity assays (Olink Target Inflammation and the newly released Olink Reveal) on 80 individual lithium-heparin plasma samples from two clinical cohorts, for a total of 1525 injections. To our knowledge, this is the first study to include Olink Reveal in a systematic cross-platform comparison and to directly compare the Orbitrap Astral with the previous-generation Orbitrap Exploris 480 under identical conditions. In our hands, the Astral delivered ~2.5-fold more identifications than the Exploris, and Seer on the Astral reached 6003 proteins (~70% of the HPPP catalogue) at near-complete albumin removal. No single protein was detected by every MS and Olink platform simultaneously, and only the deepest workflows (Seer and Olink Reveal) recovered aging-relevant pathways in a centenarian vs. adult comparison, showing that analytical depth directly determines biological interpretability. We further demonstrate that Olink Reveal can be fully automated on the Firefly System (R2 = 0.958 vs. manual preparation), enabling the throughput required for large clinical cohorts. Together, these results provide a concrete, evidence-based framework for workflow selection and support the co-deployment of deep MS and high-plex affinity assays as a comprehensive plasma-proteome-profiling strategy for clinical and translational proteomics. In this context, the present work provides a systematic and integrated benchmarking of multiple contemporary plasma proteomics workflows evaluated on a clinically relevant cohort. By comparing advanced mass spectrometry strategies and high-plex affinity-based assays under harmonized analytical conditions, this study goes beyond isolated performance metrics to examine how depth of coverage, reproducibility, throughput, and platform-specific biases influence biological interpretation. Importantly, the results demonstrate that no single workflow achieves exhaustive characterization of the plasma proteome, and that methodological choices directly shape the spectrum of detectable biological signals. The demonstrated complementarity between deep, discovery-oriented mass spectrometry workflows and highly standardized affinity-based platforms has direct implications for clinical implementation. While deep MS approaches maximize proteome coverage and enable hypothesis-generating analyses, affinity-based assays provide robust, scalable, and reproducible quantification of targeted low-abundance proteins that are often of primary clinical interest. In this study, the successful automation of the Olink Reveal workflow further illustrates how affinity-based proteomics can be adapted to high-throughput, standardized pipelines, an essential requirement for large clinical cohorts and longitudinal studies. This automation step reinforces the suitability of such platforms for routine deployment, while maintaining analytical consistency and quantitative reliability. From a clinical and translational perspective, this benchmark offers a reference framework to guide the rational selection and combination of plasma proteomics workflows according to study scale, sample availability, and intended downstream applications. By clarifying the strengths and limitations of each approach, this work contributes to improving experimental design, facilitating cross-study comparability, and supporting the gradual standardization required for clinical adoption. In the longer term, these results lay the groundwork for integrative analytical pipelines and harmonized data generation strategies that may enhance the reliability and interpretability of plasma proteomics in biomarker validation, longitudinal monitoring, and future clinical research settings.
Keywords: Affinity-based proteomics; Clinical samples; Mass spectrometry; Plasma; Proteomics