• Sign In
  1. Requests

very slow performance of skyline with saved .sky file after opening

support

View Request

view list print
very slow performance of skyline with saved .sky file after opening kguehrs  2018-02-19 05:33
 
Hello Skyline team,

I use Skyline (64 bit) 4.1.11796 with a MS1 filtering approach. There are 5 Mascot .dat files and 5 Thermo .raw files (not named the same way but correctly assigned)that I copied to my local harddrive. My computer is not very highly equipped in terms of both processors and RAM and therefore I do not expect it to be superfast. Preparing the library took quite a while but after this was finished I was able to manually remove peptides or poor peaks, edit peak boundaries for re-integration and do refinement of the protein list without too much delay. I saved the project without any problem.
However, when I try to complete the editing after opening the saved .sky file another day the performance of Skyline is very slow compared with the same manipulations on the day of project creation. Switching between samples or retention time or peak area windows needs a very long time and I see in the Windows task manager that Skyline is using lots of processor capacity just for switching between replicate tiles what I cannot understand. Is there any difference in the way Skyline links the files (spectra and chromatograms) between a freshly created and an re-opened saved file?
If the problem is my computer I would be interested to learn the minimal and recommended hardware requirement for the different versions of Skyline. Perhaps, it would be a good idea to place this information on the download site of Skyline (either as text or as a link).

Did you ever hear before about the behaviour I described. Is there any suggestion how I can prevent the current situation?
 
 
Brendan MacLean responded:  2018-02-19 08:14
First, what are your system specs? What Windows OS? what CPU, and what amount of RAM?

When you look at Windows Task Manager, what does memory consumption look like? Close to 100%?

Usually, when I have seen extreme slowness, it has been because Skyline was attempting to use more memory than the system has and so ended up what we call "thrashing", that is needing to swap memory out to disk and back into memory.

I suppose if your document is very close to the limits of your system it could be possible that Skyline ends up using more memory when you reload it from disk than when you import it directly the first time. If so, that would explain the performance you are seeing.

It is not our intent to consume more memory loading a saved file, and we have done a fair amount of optimizing this case, but it still may be true, and if this document is very close to your limits, then even a small increase may be stressing your system.

Here is a page under "Tips" on the Skyline web page that explains system requirements:

https://skyline.ms/wiki/home/software/Skyline/page.view?name=Skyline%20System%20Requirements

Maybe we will update it after learning more about your case, but it recommends at least 4 GB for basic targeted experiments and higher for proteomewide or "large-scale" proteomics experiments like yours, and it describes an ideal desktop configuration with i7 processor and 16 GB of RAM.

I suspect you have a maximum of 4 GB (less?) and you are trying to do proteomewide processing.

Thanks for posting this issue on the Skyline support board.

--Brendan