Different export formats in Diva work differently in Flowjo, and Diva 8 has some changes from Diva 6 that I thought it was worth addressing…
We have Diva 6 running on our LSRs and Diva 8 on our SORP Aria. So in Diva 6, when you export as “FCS”, the file name is “Specimen_tube.fcs”, so when you load the file into Flowjo the file is identified by the filename (which is specimen and tube), and all parameters map correctly.
When you export as “Experiment” in Diva 6, the file name for the fcs files is a 5 digit number associated with the database in Diva (“34567.fcs”), and if you use these files in Flowjo the file name in the workspace is just the number (which you can get around by having the workspace display the tube name for the file so you can identify them) and if you display H and W parameters, they are switched—FSC-H is showing FSC-W and vice versa. Area is displayed normally.
NOW, in Diva 8 (not sure about 7, I didnt have it long enough to know), BOTH exporting formats (experiment or fcs) label the file “specimen_tube_###.fcs” where the number is the order of the files in the experiment layout, so you can’t easily tell the different export formats apart…
The standards for field labels for FCS 3.1 can be found here on page 10 of 34:
but this can get confusing because different software platforms can use different fields for different things, such as the $SRC field, or “source”, is used to store the “Specimen name” in Diva.
$FIL is for the filename. For EXPERIMENT exported files in BOTH Diva 6 and 8, this is the number assigned to the file in the database, which you can also see in the D>BDDatabase>BDData folder. For FCS exported files, $FIL is “Specimen_Tube name” in Diva 6, or “Specimen_Tube_###” in Diva 8. When you load a file that was exported from Diva 8 as an experiment, Flowjo will go to the $FIL field to display in the workspace, which in this case is just a number even though the name of the actual file is from the specimen and tube.
In the Flowjo preferences, workspace tab, “How should data files be named” you can choose to specify keywords instead of the default which uses the $FIL field. Choose “$SRC” (specimen) and “Tube Name” and as you choose each it will add it to the list below the drop down menu. There is an option to “clear keywords” in the pulldown menu to clear chosen keywords. That will set the “name” in flowjo to display as specimen tube instead of filename, and FCS or experiment exported files will show up the same way, but keep in mind that Experiment exported files will STILL switch H and W signals.
Exporting in Diva
So this brings us to data management and exporting in Diva. In Diva 6 on our LSRs, if the experiment isn’t that large I typically export as experiment, which creates a folder with the name of the experiment, and in that folder creates an xml file named after the experiment and then the fcs files which are all just numbered. Then I’ll export as fcs, which if done on its own will create a folder named after the experiment, but since that already exists from the experiment export, it just sticks the fcs files in that same folder. The “experiment” exported files are numbers, and the “fcs” exported files are all named Specimen_Tube. You cant do this in Diva 8 because the file names for FCS or Experiment are the same, so you have to segregate them into separate folders and keep track of which is which.
Another local core facility (thanks Laura, who also helped me figure out some of the details of what was going on) will actually set the experiments to export as zip for all new accounts, so that the folders are kept separate that way by default (Diva will remember your choices from the previous export, so directory path, zipped, and in Diva 8 whether you want FCS 3.0 or 3.1). You really only want to use the Experiment export in Diva, so zipping it is a good way to keep it apart, and when you export as FCS it will make a new, unzipped folder named after the experiment for your fcs files.
If anyone has any other solutions you would like to share, or clarifications if I’ve gotten some details wrong, please comment or email me!