(Doesn't yet work on a Mac - sorry!!)Imagene file formatting program
The program is an Excel template with macros, so you will need to select "enable macros" if asked when you start it. You can open the program from here or save it to disk where it should perform as a normal Excel template.
The program reads in your raw IMAGENE data file and either:
Opens a set of Imagene data file and exports a column-ordered list of signal measurements for upload to eg Cyber-T
Opens a set of the Excel 'duplicates' files you have created when you looked at intra-array variability, and exports a column-ordered list of signal measurements for upload to CyberT (ie Option2 but reading the data from your excel files rather than from the raw data files)
When you open the main file, you will be located on the first sheet (ERRORS). The macro will start automatically and present you with the 3 options above. If you don't want those options, just press CANCEL.
Although the Macro that performs all the data loading etc runs automatically at startup, if you have cancelled it you can run it by clicking on the "Input and Format Imagene Files" button.
Description of Sheets
Sheet1: Errors
This Sheet allows you to compare duplicate error rates across your array. You can select whether to use Mean or Median Signals and whether to subtract background in your calculation. Background subtraction can be either by spot (subtracting each individual background reading) or by sector (subtracting the median of the background values for the meta grid of 25 spots).
The error threshold level you can adjust- the default is 50% (meaning the difference between the pairs is greater than 50% of the mean of the pairs). This in effect is a numerical version of a Bland-Altmann plot (see Martoglio...)
You can see all the data that the errors are calculated from by selecting "Show Values>>". Clicking it again will hide the values. It is worth checking that these have been calculated correctly by looking at one or two gene pairs.
You can also use the Total values if you wish, as this is calculated from the SignalMean x SignalArea (this function is therefore disabled when Median is selected
Sheet2: PAIRED DATA
This is where you will find all the data imported from your raw IMAGENE data file - in pairs. You can also see which of the pairs have poor duplicates as these will be highlighted if they exceed the threshold value set on the Errors page.
Sheet3: Unpaired DATA
This is where you will find the data listed without pairing. It is calculated from the Paired Data sheet and therefore has the empty cells removed.
Sheet4: Raw Data
This is simply a copy of your IMAGENE file without the header rows.
File Export Options
Export Paired Data
If you want to skip the normalisation process we have developed, you can simply use the average mean or median signal from each experiment and upload that data to Cyber-T or Genespring.
The pairing takes place across the array using the known location of spots. You can check that pairing has taken place properly by looking at the Gene ID column of the "PairedData" sheet. Any discrepancies in the Gene names will be highlighted.
Export UNPaired Data
This exports the data in an unpaired format, i.e. each spot is treated as an independent spot - without any regard to duplicates or pairing. There is therefore no need for averaging between duplicates. This format is particularly suited to uploading to our normalisation program.
Background Subtraction
You can select whether to subtract background values from your data
The format of this is updated regularly, suggestions are welcome (email me)
The file is in ZIP format:
(Doesn't yet work on a Mac - sorry!!)
(NB: If you get a load of gobbledygook when you try and download it, make sure you ctrl-click (PC) or right-click (Mac) and choose "Save Link as" or "Download to Disk")
Sam Saidi
Clinical Research Associate