Next-generation proteomics in medical laboratories: metrologically sound quantitative protein tests vs. innovative and personalized proteome patterns

Abstract In precision medicine protein phenotypes are foreseen that combined with pattern recognition tools exhibit great potential in guaranteeing safe and accurate test results for patient management and intervention. For discovery purposes novel high-throughput technologies have emerged and this has led to a resurging interest in plasma proteomics. These next-generation proteomics technologies pursue enhanced speed and robustness of proteome readouts and have therefore been designed in combination with machine learning tools for pattern recognition to allow large-scale population studies. In the proximity extension assay two antibodies with single-stranded DNA tags are applied that hybridize upon binding to their target proteins. The second affinity-based assay relies on DNA-based aptamers with specific binding shape complementarity to their target proteins. In both assays the resulting protein-DNA barcodes are quantified through next-generation sequencing. On the one hand the ease of data analysis and integration with genomic and transcriptomic data is attractive compared to complex MS-based proteomics datasets. On the other hand MS allows an unequivocal characterization of the protein of interest and its proteoforms. Laboratory professionals now face the following protein diagnostics dilemma: either hold firm to metrologically sound protein or proteoform quantification by mass spectrometry or pursue individual protein profiles combined with machine learning algorithms without detailing various proteoforms. The first strategy has proven slow and may not be sustainable for all biomarkers, whereas the latter approach clearly requires concerted efforts from all stakeholders involved in medical test development before these can be adapted into medical laboratories.