Hello,
question 1: In the latest skyline version, the mascot threshold for creating peptide library has changed from identity to homology. For the peptidome analysis in my lab, we wish to use identity threshold rather than homology threshold. is there anyway to choose one over the other?
question 2: when we import mascot search result into skyline, creating protein and peptide list, it seems that one peptide can be assigned to several protein isotypes, and the number of proteins ends up much more than expected. Is skyline capable of grouping proteins, or only showing the most possible proteins?
Thank you very much!
Yiying |
| |
| Skyline Support responded: |
2015-01-14 14:05 |
Hi Yiying,
1. We made this change based on feedback from other users, and also the Matrix Science team. We were under the impression that we had probably made a mistake in choosing "identity" in the first place, rather than "homology". Can you please clarify why you feel the former is more appropriate in your case so that we can again check with Matrix Science to see what they think, and whether we really need to implement the ability to choose between these scores, or whether there really is a "right" score to be using for our library building.
2. Skyline makes no attempt to group proteins. It does have some facilities to help you understand the uniqueness of your peptides within a background proteome (or FASTA file). If you really want to measure proteins quantitatively (and not just petpides), then you need to make extra effort to either use only unique peptides or make sure you really understand that other proteins that impact a the quantity of a target peptide are not actually present (e.g. through some sample prep). Protein grouping is more of protein identification concept when you are trying to estimate the probability that your sample contains a protein, and less something you should rely on for determining protein quantity in multi-replicate experiments.
Hope this helps. Thanks for using Skyline.
--Brendan |
| |
| YZ responded: |
2015-01-18 21:24 |
Dear Brendan,
Thank you very much for your response!
1. For identification of our endogenous peptides, using homology threshold ends up a ridiculously larger number of IDs. We just can't believe they are true by manually inspection of spectra. We wish that only confidently identified peptides are in the list and further quantified when importing the search result. That's why we prefer "identity".
2. I understand that it is like proteinprophet in TPP which assigns a probability score to the protein and can take spectra counting results to quantify the protein. I wish that there are some follow-up software that can take skyline quantification result and further refine the protein IDs.
Best,
Yiying |
| |
|
|