Another big thanks to @PastelBio for something I would have missed!
Okay -- so what if you took one of the remaining limitations of spectral libraries and threw it out the window? I'm talking about the fact that your library must contain the modification that you're looking for(!?!) -- Then you'd have ANN-Solo! You can read about it at JPR here. An earlier version of the text was released at BiorXiV as well.
Now...I'm unclear how the ANN (Approximate Nearest Neighbor) part of this differs from the NIST Open Search functionality added to MSPepSearch last year. At first it seems interesting that the authors use the NIST library here but don't appear to compare their code to MSPepSearch + HybridSearch. They do use other libraries and since MSPepSearch only utilizes NIST library format, maybe the comparison isn't possible? I would be very interested in seeing a comparison between the two.
Unfortunately, while Open Search has an .Exe that I can run and use, ANN-Solo requires a NumPy Python to work and I'll have to ask for help if I want to try it. Honestly, with results as good as the paper reports -- 100% worth it.
Interested and don't feel like reading on a Friday? You can get the software here!
No comments:
Post a Comment