RESUMEN
2-D gel electrophoresis has been used for more than three decades to study the protein complement of organisms, tissues, and cells. Three issues are holding back large-scale proteomics studies: low-throughput, high technical variation, and study designs lacking statistical power. We identified image analysis as the central factor connecting these three issues. By developing an improved image analysis workflow we shortened project timelines, decreased technical variation, and thus enabled large-scale proteomics studies that are statistically powered. Rather than detecting protein spots on each gel image and matching spots across gel images, the improved workflow is based on aligning images first, then creating a consensus spot pattern and finally propagating the consensus spot pattern to all gel images for quantitation. This results in a data table without gaps. As an example we show here a study aimed at discovering circulating biomarkers for chronic obstructive pulmonary disease (COPD). Eight candidate biomarkers were identified by comparing plasma from 24 smokers with COPD and 24 smokers without COPD. Among the candidates are proteins such as plasma retinal-binding protein (RETB) and fibrinogen that had previously been linked to the disease and are frequently monitored in COPD patients, as well as other proteins such as apolipoprotein E (ApoE), inter-alpha-trypsininhibitor heavy chain H4 (ITIH4), and glutathione peroxidase.
