Prosit is a deep learning algorithm developed by the Wilhelm and Kuster group to predict fragmentation and retention times for peptides. We use Prosit to predict DIA-specific libraries from all possible +2H and +3H peptides in FASTA databases. While the search time is roughly the same as FASTA searches, Scaffold DIA is specifically designed to perform proteome-scale searches, such that Prosit libraries perform as well as smaller sample or tissue-specific DDA libraries used by other tools without the need to perform any DDA library-generation experiments. This approach is particularly powerful when combined with the chromatogram library method to calibrate Prosit libraries for each sample type and instrument.
Proteome Software has built a repository of spectral libraries for common organisms which are available below. If you would like us to make a specific Prosit library for you, please contact us with an organism-specific FASTA (with only reviewed entries). Please note that Prosit will only work with carbamidomethylated peptides (C+57). You cannot use any other cysteine alkylation method and you cannot add any PTMs for Prosit other than oxidation.
See this paper for more information.
Library Creation Parameters
The following parameters were used in the conversion process and thus should match your instrumentation settings. If they vary dramatically from your settings please contact us.
|Charge Range||2 - 3|
|Maximum Missed Cleavages||1|
|m/z Range||396.4 - 1002.7|