February 18, 2026
OncoHost Publishes Landmark Study in JPBA Demonstrating Serum-Plasma Proteomic Bridging, Expanding Biomarker Research Capabilities
Original source here
BINYAMINA, Israel and CARY, N.C., Feb. 18, 2026 /PRNewswire/ -- OncoHost, a technology company transforming precision oncology through proteomics and AI, today announced the publication of a new study in the Journal of Pharmaceutical and Biomedical Analysis (JPBA), titled "Bridging the Gap: A Systematic Approach to Integrating Serum and Plasma Proteomic Datasets for Biomarker Studies."
The study introduces a validated computational framework that harmonizes serum and plasma proteomic datasets—specimen types historically considered analytically non-comparable due to biological and pre-analytical differences.
By overcoming this long-standing barrier and technical divide, the framework allows researchers to combine heterogeneous cohorts, accelerating biomarker discovery and validation efforts. The approach enhances both the analytical depth and translational applicability of proteomic research while creating new flexibility in sample utilization.
The multi-institutional collaboration included leading researchers from the National Cancer Institute (NCI), Yale School of Medicine, Heidelberg University Hospital, and biomarker development company ions.bio, alongside the OncoHost scientific team.
Unlocking Cross-Specimen Data Integration
Serum and plasma are widely used in clinical research and biobanking; however, differences in sample preparation result in proteomic variations that have limited direct comparison and data integration across specimen types. Consequently, large retrospective cohorts derived from each specimen type have often been analyzed separately.
In this study, researchers performed high-throughput proteomic profiling of 7,289 proteins using the SomaScan® platform on 177 matched serum–plasma sample pairs from cancer patients across three independent cohorts. Remarkably, 91.6% of proteins demonstrated statistically significant correlation (p < 0.05) between serum and plasma. Using matched sample pairs, the team derived linear scaling factors that were consistent across cohorts, supporting the generalizability and robustness of the bridging methodology.
"This work addresses a longstanding challenge in proteomic biomarker research," said Coren Lahav, MSc, Senior Data Scientist and lead author of the study. "By enabling systematic transformation between serum and plasma measurements, we can leverage previously siloed datasets. This strengthens analytical robustness, improves clinical sample flexibility, and supports scalable multi-cohort validation."
Validation Through PROphet®
Critically, the study validated the bridging strategy using OncoHost's PROphet® platform—an AI-powered, plasma proteomics-based model designed to predict immunotherapy outcomes in patients with non-small cell lung cancer (NSCLC).
When applied to scaled serum proteomic measurements, the PROphet model maintained predictive accuracy. Clinical benefit classification and survival stratification derived from transformed serum data were comparable to those generated from plasma, confirming preservation of the underlying biological signal.
"This milestone advances the field of liquid proteomics," said Ofer Sharon, MD, CEO of OncoHost. "Demonstrating that plasma-based predictive models can be reliably extended to serum through rigorous harmonization reinforces the robustness of our approach and supports broader clinical implementation. The ability to generalize across specimen types enhances the utility of proteomic diagnostics and supports wider adoption in precision oncology."
Broadening Technical and Clinical Potential
Beyond immediate dataset alignment, the study provides a structured methodology for future sample-type standardization, including compatibility across tube types and biological matrices. Enabling cross-specimen comparability may unlock valuable serum-based cohorts for discovery and validation initiatives.
As proteomics become increasingly central to precision oncology, systematic standardization approaches such as this will be essential to advancing reproducible and scalable biomarker development.
Link to study: https://doi.org/10.1016/j.jpba.2026.117421
