Just exploring the program a little and thought of a couple of points that might be nice.

- When importing a data file, could you enable crtl or shift selection, so multiple files can be imported from the same directory at the same time.

- If you get the mouse-over comment and use up and down to scroll between 2 data files then the comment does not automatically update. If it did then it would be much quicker to see if a particular spot was the same gene.

- Related to this, is it possible to have all genes represented in a pathway map with the genes not in your file just blanked out. I realise you could do this by giving them all a 0 in the data file, but this would take time to arrange each time, make them white rather than background colour and create much bigger data files.
The idea for this option is because although its good to see which pathways are changing its not possible to tell if they are the same genes when making a comparison, whereas if they had a 'fixed' position you could easily visualise changes of a particular gene between 2 data files.

Cheers
Matt

don't know if I understand you right;
If you switch between 2 or more data files the same box represents always the same gene.
It is not possible to blank the abscence calls in your data file, because the whole matrix gets mixed up.
Svenja

It seems this is already implemented to some extent, but is not described (or I didn't see it) in the help section.

I seem to remember from talking to oli a long time ago something about putting x's in the missing values of the data file, or do you leave them blank? I guess this then means that all genes are in the same order. However to the newuser the obvious thing seems to be to put your ID's of interest into the data file, which results in the mixing up that I was refering to.

It is quite difficult to keep all the ID's throughout an analysis, and replacing them later requires time and knowledge of using something like access. Surely MapMan 'knows' what ID's are present on the affy chip (or in any mapping file) and so wouldn't it be useful to set the default to treat the ID in the same way as if it has a 'x'. ie: if the ID is not in the datafile it's treated as an 'x'.

Also what would happen if the data file ID's are sorted in a different order? I assume position is defined wholely by ID and that the order in the datafile has no effect.

At the least a line in the data format section of help such as 'Identifers that have no measured value are indicated by the value 'x'.' would have prevented my mistake.

Sorry for the confusion. ???

I ) Use "add experiment directory", it is on the right mouse button when you click on the experiments folder.
It will read a whole directory and will interprete all files as data files.
II) We need to have the all "genes" (features) in all the data files you want to compare. The software could automatically blank out ('X') all missing values, that would result in lots of unknown unknown values.
As standard, only values which are in the data file (present or absent) are displayed. This helps keeping the overview.