BMC Bioinformatics

official impact factor 3.03

Open Access Highly Access Methodology article

Correlation analysis of two-dimensional gel electrophoretic protein patterns and biological variables

Werner Van Belle1*, Nina Ånensen2, Ingvild Haaland2, Øystein Bruserud2,4, Kjell-Arild Høgda3 and Bjørn Tore Gjertsen2,4*

Author Affiliations

1 Bioinformatics Group, Norut IT, Research Park Tromsø, Postboks 6434, N9294 Tromsø, NO, Norway

2 lnstitute of Medicine, Hematology Section University of Bergen, Bergen, NO, Norway

3 Earth Observation Group, Norut IT, Research Park Tromsø, Postboks 6434, N9294 Troms0, NO, Norway

4 Department of Internal Medicine, Hematology Section Haukeland University Hospital, Bergen, NO, Norway

For all author emails, please log on.

BMC Bioinformatics 2006, 7:198 doi:10.1186/1471-2105-7-198

Published: 10 April 2006

Additional files

Additional file 1:

Animated Correlation Method. A movie illustrating the correlation method made with blender [61]. The movie can be played with mplayer [62].

Format: AVI Size: 4.6MB Download file

Playing the movie within this page requires QuickTime 3 or later and JavaScript. Read more

Open Data

Additional file 3:

Correlation images illustrating impact of misclassification. The impact of wrong ALL versus AML diagnosis was examined by random swapping ALL and AML labels in the AML/ALL versus 2DE image correlations. This results in lower correlation values as expected.

Format: PNG Size: 381KB Download file

Open Data

Additional file 2:

Spot shapes. Two images showing the difference in spot sizes between M0/M1/M2 and M4/M5 samples. The process of the changing spot distribution can be visualized by sorting all images according to their FAB classification and then showing them chronologically. This is visualized in a small movie. The two images and the movie are contained within a zip file. It can be extracted using unzip [63,64]. The movie can be played with mplayer [62].

Format: ZIP Size: 981KB Download file or display content in a new window

Open Data

Additional file 4:

Alignment accuracy. The correlation algorithm relies on the correct alignment of input images. This process is typically performed using calibration marks on gels. The additional data illustrates the importance of correct alignment.

Format: PNG Size: 1.4MB Download file

Open Data

Additional file 5:

Source material. The source material includes correlation analysis, image coloring, Gaussian bumps and the simulated images. The source is contained within a .html file. The algorithm is implemented in IDLv6.1 [60].

Format: HTML Size: 15KB Download file

Open Data