SMSNet's trained models

Two models for SMSNet are provided here. The "m" model was trained on WCU-MS-M dataset which contains 22.6 million MS/MS spectra acquired on Q Exactive mass spectrometers at the Proteomics and Metabolomics Core Facility, The Wistar Institute. This model allows SMSNet to predict 20 unmodified amino acids and oxidized form of Methionine. The "p" model was trained on WCU-MS-P dataset which contains 24.5 million MS/MS spectra (22.6 million in WCU-MS-M plus additional spectra of phosphorylated peptides). This model allows SMSNet to predict 20 unmodified amino acids, oxidized Methionine, and phosphorylated Serine, Threonine, and Tyrosine.