Premessa-output files-FlowJo naming problems about premessa HOT 16 CLOSED

The parameter name thing you mention is on purpose. When the $PxS parameter is missing, the software will copy the $PxN value in the $PxS. I think this is the best approach because there are a number of downstream analysis programs that have problems when the $PxS is missing (citrus was one of them for instance, although this has been fixed now). Unless this is creating serious problems, I'd rather keep it that way.

With regards to the file names. You mean that the actual FCS files have different names on disk, but appear all with the same name when loaded in FlowJo? If that's the case I suspect it may be due to the fact that I am not updating one of the FCS keywords and FlowJo is looking in that keyword to determine the file name, instead of taking the file name on disk. Can you put up on a shared folder a couple of those FCS files so that I can double check?

from premessa.

This is probably due to the $FIL keyword not being updated. You can check it in R with the following:

library(flowCore)
fcs <- read.FCS('/path/to/your/file')
description(fcs)['$FIL']

from premessa.

Should be fixed in the latest commit. @mleipold can you please reinstall the software and test it?

from premessa.

Hi Federico,

I went through the entire reinstall process for Premessa (starting at install.packages("devtools")). It does not seem to be fixed.

I took the same two original *.FCS files (directly off Helios; note uppercase), and did the Premessa Normalization on them. This gave me two *_normalized.fcs files (note lowercase).

I then put the two originals and the two daughters into Mac FJ10.1.r5.

The "Name" is still coming up as *.FCS (uppercase) with no "_normalized". If I add a column and look at "File Name", then "File Name" lists the correct file naming (pairs of *.FCS and *_normalized.fcs)

In other words, looking only at "Name" (the FJ default), there's no way to tell original from daughter.

I have attached a screenshot, including showing that Finder is listing the daughter files correctly as "*_normalized.fcs".

from premessa.

Hi Mike,

Sorry about that. I don't have access to a copy of FlowJo so testing that is not easy for me.

It should be fixed in the latest version. Can you please install it again and test? By the way to reinstall the package you only need these lines

library(devtools)
install_github("ParkerICI/premessa")

from premessa.

Hi Federico, @pfgherardini

Here's a Drive folder with some files:
https://drive.google.com/open?id=0Bz9sd8nC272jdHVhNnFaQ0dFTE0

For the comparison within FJ, I only looked at the "2x" file (original vs norm'd).

To confirm: you want me to install it a second time today (I already reinstalled it once today, from your request last week)?

from premessa.

Yes I was asking if you could reinstall again today and try to normalize the files again. I submitted a fix ~20 minutes ago that should fix the problem. Let me know.... if it still doesn't work I'll try and get my hands on a copy of FlowJo

from premessa.

@pfgherardini

OK, did the reinstall a second time, just using the
library(devtools)
install_github("ParkerICI/premessa")

part as you stated.

Unfortunately, it's still not working.....same Name/File Name issue.

I doubt this has anything to do with it, but I'm running R 3.3.1, and RStudio Version 1.0.153. I notice during the RStudio execution, it states that 'shiny' and 'devtools' were written under R 3.3.2.

Since I can get the program to execute, I doubt that's a problem. But, in the interest of troubleshooting, I wanted to mention it just in case.

from premessa.

Ok let me look into this more. Thanks for mentioning the version differences but I don't think those represent the problem.

from premessa.

from premessa.

That's what I did but it didn't seem to fix the problem. At least on my machine, the old file name is nowhere to be found after it goes through the normalizer

from premessa.

from premessa.

Alright should be fixed for real now! :)

@mleipold can you please install again and retry? Apologize for the back and forth....

from premessa.

Hooray, it appears to be fixed! Thanks!

We've been discussing it only at the Normalizer level; does this propagate to the Debarcoder as well, or would that be a separate fix, if necessary?

from premessa.

The fix will work across the board, so for the normalizer, debarcoder, and also FCS panel editor

from premessa.

Great. If I find some issue with it down the road, I'll add it to this thread.

But otherwise, I consider this closed.

from premessa.