MCGH: Analysing microarray-based CGH experiments

Junbai Wang*, Leonardo A. Meza-Zepeda *, Stine Henrichson Kresse*, and Ola Myklebost*

Department of Tumour Biology*, Norwegian Radium Hospital, N0310 Oslo, Norway

MCGH is a Matlab toolbox with graphical interface to analysis DNA microarray-based comparative genomic hybridization (CGH) experiments. CGH on DNA microarray enable us to measure the copy numbers of thousands of genes through the genome. M-CGH allows us to visualize the copy number changes (amplifications or deletions) and rapidly identify the amplification or deletion regions by directly loading the export file from microarray image analysis package, such as GenePix and QuantArray.

MCGH screen shot
MCGH Screen shot MCGH main window MCGH Help MCGH Plot 1 MCGH_Plot 2 MCGH_Plot 3

You can download the MCGH toolbox: mcgh.tar.gz

Please read "install.txt" and license file before installing the Matlab toolbox.
Feel free to use the package in your academic research. If you use it, please cite the following paper:

Wang J, Meza-Zepeda LA, Kresse SH, Myklebost O.
M-CGH: analysing microarray-based CGH experiments.
BMC Bioinformatics. 2004 Jun 9;5:74.

1) 7th of July 2004: New version of M-CGH was updated. Some bugs in the early versions were fixed at here!

2) 15th of July 2004: Some notes about how to use M-CGH were added in the M-CGH help menue

3) Jan. 2005: One bug in the new Windows version of MATLAB. If you can not plot individual Chroms. in the M-CGH please replace "break" command as "return" in file "mcgh_ChrPlot.m".

4) Feb. 2005: Some minor bugs in "MCGH property" window have been fixed. Please download the toolbox and replace the old files with new ones in your MATLAB. Thanks the bug report from Jessica J. Connelly in Duke university!

5) Apr. 2005: Now, MCGH can handle the input data with single spot if there is no duplicated clones.

6) May. 2005: Added a pop-up menu to choose ratio=532/635 or 635/532 when the input data is GenePix format.

7) May. 2005: Added a color list which will label spots as different color according to their pre-defined classes, a sample color list is list_of_three_classes.csv . Currently, MCGH only support five different types of colors!

8) May. 2005: Added a list of information about missing spots, i.e. the number of spots which are flaged, filtered and no clone location information will be displayed in the main window of MATLAB.

9) May. 2005: For Amplicon boundary estimation, Wavelet approach is faster and better than fuzzy knn method when the input data is large.

10) June. 2005: For "M-CGH Property", added N*Background Intensity in "Signal to noise level".

11) July. 2005: A problem of convergence for certain dataset is solved in MCGH.

12) August. 2005: A manual adjustion of ratio (reference vs. control sample / constant value) is added in MCGH.

13) Oct. 2005: Improvement was made in loading and exporting data.

14) Nov. 2005: Some features of MCGH. All plot and figures in MCGH are interactive that means a red circle will be marked when you click a spot on the figure. You can remove such circle by clicking the button 'Refesh figure of raw ratio plot' or 'Genome plot with mean centered Ratio'. The information of red circled spot will be shown at the rightdown corner of MCGH main window. The 'M-CGH find clone' function can help you locate the clone that you are interested in the figure by inputing its name.

15) Sep. 2006: The data input function was improved. Now, MCGH is able to load GenePix file that contains the empty cell.

16) Nov. 2006: An easy way to add the M-CGH toolbox to the MATLAB search path, please copy below script use_mcgh.m to your MCGH folder then type use_mcgh(1) for adding path or use_mcgh(0) for removing path in the MATLAB main window.

To run the program please read the short description of how to use this toolbox in "M-CGH help" menu or the "readme.txt". Comments and questions about the toolbox please send email to Junbai Wang ( )

M-CGH is an open soure program which means you are free to use and modify it by your own interesting. Also, you are ready to take the risk of errors and bugs inside program.

Free counter and web stats