The identification of biomarkers has been one of the major applications in proteomics research. The high specificity and sensitivity of a good biomarker can be evaluated only in population studies, which is costly and a limited number of candidate proteins can be evaluated. However, high throughput proteomics usually identifies thousands of proteins from a biological sample and yields tens to hundreds of candidate biomarkers. We demonstrate the use of robust and flexible spectral counting, a label-free semiquantitative method, on relatively abundant proteins to integrate proteomic results from different samples obtained from different laboratories for reliable biological insights and high-quality biomarkers. Furthermore, this strategy can be easily tuned for one or a few biomarkers on demand. We illustrate the use of this strategy to reveal the biological differences between terminally differentiated myocardium and rapidly proliferating HEK293T cells, as well as the successful selection of both positive and negative maturation markers for cardiomyocytes differentiated from human induced pluripotent stem cells (hiPSCs). The advantage of this strategy is the capacity to utilize the exponentially growing number of proteomic results deposited in the public repository for candidate biomarker selection and to derive biological insights.
Chang et al. (Mon,) studied this question.