Protein 2D Gel Electrophoresis Images Matching with Maximum Relation Spanning Tree

The electrophoresis chromatography, a popular analysis tool, is able to separate different kinds of protein pro£les. In this paper, we present a robust comparative algorithm Maximum Relation Spanning Tree (MRST) for matching large scale and large sets of two dimensional protein gel electrophoresis (2DGE) chromatography images without the need of a priori landmark. The algorithm not only can handle the conditions of image rotation, shift and reverse, but also can handle fractional mapping problem. In the matching process, we apply fuzzy inference technique to conclude the £nal decision of mapping and location. The proposed system presents up to 94% correct matching performance for 225 2D gel test images. The additive value is the foundation of comparing small gel images to large format gel images and the constitution of searching scheme for a huge two dimensional gel electrophoresis chromatography image database.