SMSNet's training datasets (WCU-MS-ALL)

The full WCU-MS-ALL training datasets for SMSNet that was used to train the released model (DOI: 10.6084/m9.figshare.8259122). This dataset contains roughly 25 million MS/MS spectra of unmodified peptides, peptides with Methionine oxidation, and peptides with phosphorylation. Identifications were made with MaxQuant (various versions) with FDR &lt;5% at peptide level.<br>The file format is tab-separated. The columns contain: peptide sequence, precursor charge state, observed precursor m/z, PEP score (from MaxQuant), Andromeda search score (from MaxQuant), and observed MS/MS peaks. The dataset is split into 3 files due to 5GB size limit.