[Bioc-devel] FCS3.0 DiVa files?

Hi Byron,
do you think you could send me one of the diva files for testing
purposes. I like to test against as many different file types as
possible whenever I do changes to the import function and my supply is
somehow limited to FACS Array and FACS Calibur files. By the way, FSC
and SSC (and sometimes also other parameters) are viewed linearly
because the amplification of the measured signals by the instrument is
either linear or logarithmic, this can be set by the user. You can get
the type of amplification for each parameter n from the $PnE keyword in
the FCS file.
Cheers,
Florian

Byron Ellis schrieb:

Actually, I stand corrected---I was looking at untransformed data and
comparing to transformed data (the Forward Scatter is usually viewed
linearly for whatever reason so it looked as it should). It looks like
the SVN version (as of today) reads the FACSDiVa files correctly.
Hooray. For the longest time it looked like there was an alignment
problem or something, but apparently not.

Of course there are negative entries for values that are,
theoretically, photon counts but that's more the fault of the
equipment manufacturers who thought that hardware background
correction was a good idea. :-)

P.S. Turning off a dump of the data matrix by default was a fantastic
idea. My test data is 13x~520K  and that just kills Aqua R.

On 9/5/06, Wolfgang Huber <huber at ebi.ac.uk> wrote:

Hi Byron,

would it be possible to be more specific than "weird sorts of systematic
errors", and for instance give some example files together with what you
consider correct results?

   Best wishes
   Wolfgang



Byron Ellis wrote:

The latest readFCS unfortunately doesn't produce sensible results
except for the first parameter (generally forward scatter), the other
parameters have weird sorts of systematic errors that almost look like
alignment problems. At least two other people in my lab here have Java
importers for these sorts of files that I'm considering to be
"correct" and working from there.

On 9/4/06, Florian Hahne <f.hahne at dkfz-heidelberg.de> wrote:

Hi Byron,
As far as I know these files don't seem to adhere to the

specifications

in the FCS 3.0 format. I had one example file which I read in using a
number of different software tools and almost all of them produced
different results.
readFCS could read the file without errors and the plots of the data
looked ok. The only problem is that the absolute values are not

the same

for the different software, but the measurement relations are the

same,

i.e. the data is on different scales. And since I don't know what the
correct values are (and are there "correct" values at all? In the end
these are just arbitrary dimensionless numbers that somehow

reflect the

fluorescence intensities...) I didn't know how to proceed.
Please feel free to take a stab at the import if you think that is a
problem or send me one or two of your files along with some info

about

the expected range so I can give it another try.
Cheers,
Florian

Byron Ellis schrieb:

Hi all,

I remember a couple of months ago there was a brief discussion about
FCS3.0 files, in particular DiVA files recorded in floating

point. Was

there ever any progress made on reading those in? I took a look
through the logs, but I may have missed something. I suddenly find
myself working with an LSR2 using DiVa that generates a lot of these
files. :-)

If not, I'll take a stab at it since I need to read these files

sooner

rather than later.

--
Florian Hahne
Abt. Molekulare Genomanalyse (B050)
Deutsches Krebsforschungszentrum (DKFZ)
Im Neuenheimer Feld 580
D-69120 Heidelberg
phone: 0049 6221 424764
fax: 0049 6221 423454
web: www.dkfz.de/mga

--
Florian Hahne
Abt. Molekulare Genomanalyse (B050)
Deutsches Krebsforschungszentrum (DKFZ)
Im Neuenheimer Feld 580
D-69120 Heidelberg
phone: 0049 6221 424764
fax: 0049 6221 423454
web: www.dkfz.de/mga