Biomarkers, quantitatively measurable indicators of biological or pathogenic processes, once validated play a critical role in disease diagnostics, the prediction of disease progression, and/or monitoring of the response to treatment. They may also represent drug targets. A number of different methods can be used for biomarker discovery and validation, including proteomics methods, metabolomics, imaging, and genome wide association studies (GWASs) and can be analysed using receiver operating characteristic (ROC) plots. The relative utility of single biomarkers compared to biomarker panels is discussed, along with paradigms for biomarker development, the latter in the context of three large-scale biomarker consortia, the Critical Path Predictive Safety Testing Consortium (PSTC), the NCI Early Detection Research Network (EDRN) and the Alzheimer's Disease Neuroimaging Initiative (ADNI). The importance of systematic optimization of many parameters in biomarker analysis, including validation, reproducibility, study design, statistical analysis and avoidance of bias are critical features used by these consortia. Problems including introduction of bias into study designs, data reporting or data analysis are also reviewed.
Keywords: Biomarkers; Disease diagnosis biomarker discovery; Polypharmacy; Protein biomarkers.
Copyright © 2013 Elsevier Inc. All rights reserved.