Re: Using DIVA Exp Files in Flojo

On 11/10/2007, at 2:16 AM, Houston, Jim wrote:

> We have been using the BD Diva Files for some time in FloJo.	I now  
> have another problem.
>
> I know that DIVA uses the Specimen name when exporting files as  
> FCS.	We usually use the same Exp name and Specimen to make things  
> less complicated.
>
> Now I have a set of experiments that we run with different specimen  
> names under 1 exp.  In FloJo you cannot seem to access the Specimen  
> name directly.  FloJo support was unable to help in this also.  BD  
> support is slow at best.
>
>
>
> If you have experience in this area please let me know some clues  
> on how to handle this.
>
>
>
> I have also notice using the DIVA software that when you change  
> some user keywords they do not export out.  Any idea with what is  
> up with that?



Hi Jim,

(this is my understanding of how it works - if anyone can add or  
change anything I would love to hear it!)

Diva can export data in three different ways:-

- FCS2 - I don't know what it does with keywords here as I never use it
- FCS3 - FCS3 files are rewritten and renamed during the export  
procedure
- Experiment - FCS files are moved out of the database along with an  
xml file.

In both FCS3 and Experiment exports it looks like the specimen name  
is written to the "$SRC" keyword. You can use FlowJo preferences to  
use that as part of your workspace display name and/or you can add it  
is a column in the workspace display.

In the case of an Experiment export you will certainly want to check  
the preferences for display name or you may end up displaying the  
filename which in this case is just a number (assigned by the Diva  
database). In an FCS3 export the filename is actually created from a  
concatenation of the Specimen and Tube names.

When you are exporting as Experiment any changes in keywording made  
after the files are recorded do not show up in the FCS files  
themselves because the files are just moved, not rewritten. The  
changes are included in the accompanying xml files (which is not  
(currently?) much use in FlowJo).  If you export as FCS 3 the changes  
in keywords are included in the files as they are rewritten.

Because of the keyword issue we have been exporting as FCS3 files but  
that creates two problems - (i) it is much slower and (ii) the	
filenames can end up longer than 28 character which can create	
problems keeping track of them in FlowJo (Mac) if you move them  
around after starting your analysis.

What would be ideal would be an export mechanism that resulted in  
updated keywords in FCS files but also retained the xml (so Diva  
experiment specific information can be retained in case it needs to  
be reimported into Diva). Ideally this could be run as a batch job  
overnight in order to maximise analysis time during working hours.  
Anyone in the software group at BD listening :)

Regards,

Adrian Smith
Centenary Institute, Sydney, Australia