Format of input data

The format of data is shown in file ChipLog1.txt, Example05.txt in the Sample folder.

The first two columns in the above data file represent geneID and arrayID. The third and forth columns represent two experimental factors. If there is only one experimental factor, please put the experimental factor in the third column and set the forth column are all one. The fifth column represents the dye (one represent cy3, two represent cy5). If you have no dye-swap design, just set dye values are all one. The last column is Expression levels. Missing data is denoted by 99999.000.

This model is developed by Yan Lu.

This is Coefficient-setting Box of Microarray, which will pop out automatically when you selected Microarray menu.

　

1. Open File
Because microarray files are often very large, here we use a different way to open microarray files. First, click the New in the QGAStation. Click the Microarray method to open the Coefficient-setting Box. Then, open the file through button .
2. Method
There are two methods provided here, Mixed Model Approach and t-test Model. If select t-test Model. the options below are all unavailable.
3. Array Effect
If there is array effect within experiment, please choose Yes, or No without array effect.
4. Have Replication
If select this option, the program will construct conditional data (t|t-1) from the observed traits. If the option is not been chosen, the step option will not be enable. The result file will automatically save as filename.cos.
5. Experiment Factors
Select One Factor if there is only one experimental factor, and select Two Factor if there are two experimental factor. 6. Linkage Rule There are 8 Linkage Rules listed here, single linkage, complete linkage, median linkage, centroid linkage, unweighted pair-group average, weighted pair-group average, flexable-beta method, and ward's method.
7. Dye Design
If the experiment is a Dye-Swap design, please select Yes else No for a not Dye-Swap design.
8. Jackknife Kind
If the Replication option is No, the Jackknife Kind is autumatically Array. If the Replication option is Yes, Both Array and Replication are available.
9. Jackknife Number
If the Jackknife Kind is Block, Jackknife Number resampling unit is default 1. If Jackknife Kind is Cell, Jackknife Number  of resampling unit can range form 1 to 9.
10. Significance Level for Single Gene Model
There are six options of significant level for single gene model: 0.01, 0.005, 0.001, 0.0005, 0.0001 and 0.00005. For mixed model, the default value is 0.001.
11. Significance Level for GT Effect Clustering
There are four options of significant level of GT Effect for clustering: 0.05, 0.01, 0.005 and 0.001. For mixed model, the default is 0.001.
12. Run
Run the data with selected model and model.
13. Cancel
Leave the Coefficient-setting Box, and do nothing.
	Home