Warping two-dimensional electrophoresis gel images to correct for geometric distortions of the spot pattern
Journal article, 2002

A crucial step in two-dimensional gel based protein expression analysis is to match spots in different gel images that correspond to the same protein. It still requires extensive and time-consuming manual interference, although several semiautomatic techniques exist. Geometric distortion of the protein patterns inherent to the electrophoresis procedure is one of the main causes of these difficulties. An image warping method to reduce this problem is presented. A warping is a function that deforms images by mapping between image domains. The method proceeds in two steps. Firstly, a simple physicochemical model is formulated and applied for warping of each gel image to correct for what might be one of the main causes of the distortions: current leakage across the sides during the second-dimensional electrophoresis. Secondly, the images are automatically aligned by maximizing a penalized likelihood criterion. The method is applied to a set of ten gel images showing the radioactively labeled proteome of yeast Saccharomyces cerevisiae during normal and steady-state saline growth. The improvement in matching when given the warped images instead of the original ones is exemplified by a comparison within a commercially available software.

Two-dimensional gel electrophoresis

Alignment

Proteomics

Image analysis

Curved field lines

Author

John Gustafsson

Chalmers, Department of Mathematical Statistics

University of Gothenburg

Anders Blomberg

Mats Rudemo

University of Gothenburg

Chalmers, Department of Mathematical Statistics

Electrophoresis

0173-0835 (ISSN) 1522-2683 (eISSN)

Vol. 23 11 1731-1744

Subject Categories

Biochemistry and Molecular Biology

Other Physics Topics

Probability Theory and Statistics

Computer Vision and Robotics (Autonomous Systems)

DOI

10.1002/1522-2683(200206)23:11%3C1731::AID-ELPS1731%3E3.0.CO%3B2-%23

PubMed

12179995

More information

Latest update

9/12/2023