SpectroVQ Training Data, Model Weights and Evaluated Spectral Library

Supplementary data for the paper SpectroVQ: Noise-Aware Compression of Proteomics Data via Vector-Quantized Deep Learning improves MS/MS data storage and Peptide Identification. 

Both SpectroVQ's model weights (spectroVQ_weights.ckpt), the 3 species spectral libraries (splibLibraries.zip) generated with Seq2MS https://github.com/Jerryccm/Seq2MS for the evaluation of SpectroVQ-compressed data, and the full peptideAtlas data for training and internal evaluation (training_data.feather) are provided here.