Supplementary Figure 8: Effect of the refined Prosit iRT model on DIA spectral libraries.
From: Prosit: proteome-wide prediction of peptide tandem mass spectra by deep learning
Evaluation of the general (top) and refined indexed retention time (iRT) (bottom) prediction model of Prosit on C. elegans (a), E. coli (b) and S. cerevisiae (c) data from the Bruderer et. al. dataset. For refinement, the project specific HEK library was used. The required iRT window that would encompass 95% of all peptides is indicated by the red dashed lines. Sample number n and Pearson correlation are indicated.
