Importing Parameter Scan data into (GraphPad) Prism

[Jonathan Rudge]

Does anyone have any experience of importing COPASI Parameter Scan data into Prism?

I am trying to import a scan with 9 sets of values (i.e. 8 intervals) for each species.  I only want to show the results for one of these species, but cannot get the data into the XY data table correctly so that it will plot the results properly.

Looking at the data text file, I see that for each species there is a single column, which needs to be split up into 9 separate sub columns, one for each set of values.  However, getting Prism to do this correctly during the import is doing my head in.  Instead, I have to do it manually in  Excel first, before copying and pasting it into my Prism data table.  I wouldn't mind this normally, except that I have many other parameter scans that I will be wanting to graph, and it's a little time-consuming.

I have tried setting up my XY data tables with single Y values for each point, and with 9 replicate values in side-by-side subcolumns.  Neither seems to work.

It may be that I'm not putting the correct values in the Placement fields in the Import and Paste Special Choices dialogue, as I don't fully understand what they mean.  Does anyone have any experience of this?

[Pedro Mendes]

Hello Jonathan,

COPASI can output the data in very specific formats, such that you
possibly can create a report that has the format intended. However since
I never used Prism, I don't exactly know what the format is. Others here
that may have used Prism please answer as well!

> I am trying to import a scan with 9 sets of values (i.e. 8 intervals)
> for each species.  I only want to show the results for one of these
> species, but cannot get the data into the XY data table correctly so
> that it will plot the results properly.

A scan of species? presumably their initial values?

> Looking at the data text file, I see that for each species there is a
> single column, which needs to be split up into 9 separate sub columns,
> one for each set of values.  However, getting Prism to do this correctly
> during the import is doing my head in.

I don't understand why you have to create 9 sub-columns (what are
sub-columns anyway?). Maybe if you describe the input format I may be
able to help you defining a report format that would suit what you want.

When doing parameter scan, COPASI outputs one line per iteration of the
scan. Then in each column it can output basically any variable or
parameter. I assume you are using one of the report formats created with
the "Output Assistant"? Instead you can also create your own report and
put anything there. I can help you with that but it would be better if
you could say what parameters are you scanning, which operation you are
scannign (eg time course, steady state) etc.

> I have tried setting up my XY data tables with single Y values for each
> point, and with 9 replicate values in side-by-side subcolumns.  Neither
> seems to work.
>
> It may be that I'm not putting the correct values in the Placement
> fields in the Import and Paste Special Choices dialogue, as I don't
> fully understand what they mean.  Does anyone have any experience of this?

These questions seem to be Prism specific so I am unable to help there..

best wishes
Pedro

[Jonathan Rudge]

Okay, sorry Pedro.  I probably haven't use the correct terminology.  But, to give you an easier example, I've attached a text file, with the output from a parameter scan on initial particle numbers for a single species (talin).  There are 4 such initial particle values: 100, 1000, 10,000 and 100,000. 

I am interested in only one output value from the scan, a global assignment variable called Fibronectin-ligated integrin species (F-LIS). I've also attached an Excel worksheet, showing the desired output.  (Here, I've used the column title Total FAs to denote F-LIS.) In order to obtain this output I had to take one continuous column for F-LIS from the data file and break it down into the 4 "subcolumns", one for each initial talin particle value. (The worksheet also shows the desired plot.)

Rather than laboriously manipulating the data in Excel files every time and then copying and pasting it into a Prism data table, I would much prefer to be able to import the original data text file directly into it Prism, using the various transform options in the Import and Paste Special Choices dialog to automatically convert it into the correct format.  I.e., as in the Excel worksheet.

I hope this is much clearer.

Best wishes, Jonathan

[Jonathan Rudge]

Sorry, I forgot to say that I am scanning a time course, using the Output Assistant Particle Numbers, Volumes, and Global Values format.

On Tuesday, 31 March 2020 12:38:29 UTC+1, Jonathan Rudge wrote:

[Hoops, Stefan (sh9cq)]

Hello Johnathan,

Since you are only interested in Fibronectin I would modify the report
to only include the Time an Fibronectin columns.

Unfortunately the individual time courses cannot be written side by
side since they are generated consecutively.

My limited experience with PRISM is that the data selection for plots
is very similar to Excel and thus you should be able to define the
different curves of your plot independently from each other. This
approach should even work with the original output file.

Thanks,
Stefan


On Wed, 2020-04-01 at 07:39 -0700, 'Jonathan Rudge' via COPASI User

--
Stefan Hoops, Ph.D.
Research Associate Professor
Biocomplexity Institute & Initiative
University of Virginia
995 Research Park Boulevard
Charlottesville, VA 22911

Phone: +1 540 570 1301
Email: sho...@virginia.edu

[Jonathan Rudge]

Thanks, Stefan.  Simplifying the report that way has made it much easier to understand what's going on when importing the data text file directly into a Prism data table.  I can now do these imports reliably, so that only some basic editing (removing some redundant columns and one redundant row) is required before I can plot graphs directly from the table.

This should save me a lot of time in the future!

Thanks again.

Jonathan