Current Technological Challenges in Biomarker Discovery and Validation

Abstract

In this review we will give an overview of the issues related to biomarker discovery studies with a focus on liquid chromatography-mass spectrometry (LC-MS) methods. Biomarker discovery is based on a close collaboration between clinicians, analytical scientists and chemometritians/statisticians. It is critical to define the final purpose of a biomarker or biomarker pattern at the onset of the study and to select case and control samples accordingly. This is followed by designing the experiment, starting with the sampling strategy, sample collection, storage and separation protocols, choice and validation of the quantitative profiling platform followed by data processing, statistical analysis and validation workflows. Biomarker candidates that result after statistical validation should be submitted for further validation and, ideally, be connected to the disease mechanism after their identification. Since most discovery studies work with a relatively small number of samples, it is necessary to assess the specificity and sensitivity of a given biomarker-based assay in a larger set of independent samples, preferably analyzed at another clinical center. Targeted analytical methods of higher throughput than the original discovery method are needed at this point and LC-tandem mass spectrometry is gaining acceptance in this field. Throughout this review, we will focus on possible sources of variance and how they can be assessed and reduced in order to avoid false positives and to reduce the number of false negatives in biomarker discovery research.

Functional Roles of Protein Nitration in Acute and Chronic Liver Diseases

“…For instance, the levels of nitrated proteins in plasma, platelets, and liver tissue were positively correlated with the severity of hepatic cirrhosis [132] and diet-induced obesity [217]. In fact, an excellent review article for identifying potential biomarkers from its discovery to its application in clinical studies has been recently reported [218]. The initial stage is known as the discovery phase with identification of biomarker candidates.…”

Proteomic detection of nitroproteins as potential biomarkers for cardiovascular disease

“…The course from biomarker discovery to its clinical application can be divided into different stages [28]. The initial stage is known as the discovery phase and reveals biomarker candidates via genomic and proteomic analysis.…”

Multidimensional chromatography coupled to mass spectrometry in analysing complex proteomics samples

“…to understand the molecular mechanism of the underlying biological phenomena). This is particularly important in biomarker discovery 69–73 as well as in systems biology 74 studies. Another goal of data processing may be to optimize and assess the performance of LC n ‐MS systems and to compare them to other profiling methods such 1D‐LC‐MS or 2DGE.…”

“…For that reason, we provide an overview on the most important aspects that such data processing workflow should take into consideration for accurate processing of these types of data. Quantitative data processing workflows for label‐free 1D‐LC‐MS experiments comprise raw data smoothing, peak or feature detection, correction for mass shifts, time alignment and matching of the same peaks or features across multiple chromatograms followed by linking peak identity obtained with peptide sequence identification based on MS/MS spectra to peak quantity 70–72. Methods using stable isotope labelling contain additional modules to extract the relative quantitative peak information from different samples based on defined mass shifts or reporter ion intensities.…”