Hi Brendan,
I'm really impressed with your background correction feature. I avoided moving analysis of a project to skyline since the controls were so noisy (after treatments the signal went sky high though) but I recently did since i wanted to re-analyze my entire dataset for consistency and my manually integrated data is a lot like the skyline results. Well done! However, I was wondering if you have any more info or references for your background subtraction approach that aren't so technical that I might get something out of them. I've always been fascinated by it and would like to read a bit more.
Thanks,
Jason |
| |
| |
| maccoss responded: |
2010-10-20 12:25 |
Hi Jason,
So I'm not entirely sure what you mean by "background correction feature." We integrate and background subtract all chromatographic peaks using pretty standard approaches. The background subtracted is the area below a line drawn between the intersection of the start and stop integration lines and the chromatogram. Almost all modern peak integration routines perform this background subtraction. The algorithm we use to determine the start and stop of the integration is a little unique and we have been discussing writing up how we are doing this since we feel the approach used since v0.6 is working fairly well -- although like all integration routines isn't perfect.
I'm glad that you are finding that Skyline's peak integration is giving good results, it should be as good or better than commercial algorithms. If not, we want to hear about it. Low abundance peaks will always be hard to integrate and in many cases these are below the lower limit of quantitation anyway and while they provide useful differential metrics they won't be quantitative regardless of the integration algorithm.
We don't currently have any documentation written up the peak detection and integration algorithm used in Skyline. We have focused on getting Skyline features out to users and that means that most of what is performed in Skyline remains unpublished work. We have felt that it is more important to get users using the new features since that pushes things forward and gets us critical feedback more quickly. We are now beginning to focus on getting several manuscripts together and the description of the peak detection and integration and the validation that it is working correctly and reporting why we do things the way we do it is a priority.
I hope this helps somewhat. We'll be sure to send you documentation on who we determine the start and stop points as soon as we get it written up. There are a number of steps that we perform and it takes some figures to illustrate how it is done.
-Mike |
| |
| Jason responded: |
2010-10-20 13:05 |
Hi Mike,
I think its my lack of proper terminology that led to the confusion as I do simply mean background subtraction. I think it works wonderfully. I look forward to seeing your approach it when it comes out in press and do greatly appreciate the effort to push Skyline forward at a furious pace.
Jason |
| |
|
|
|
|
|