Avoiding spots detection in analysis of electrophoretic gel images

Analysis of proteomic data, presented in form of 2D gel electropherograms, is still a real challenge and requires further improvements. Three main problems in data analysis are associated with (1) ‘software-induced variance’ (due to the different methods of data pre-processing or their different input parameters), (2) univariate significance analysis, and (3) requirements of spots detection and quantification. In our study, the advantages of multivariate significance analysis and a pixel-based approach are demonstrated upon the example of the real data set.