Figure 6 | Scientific Reports

Figure 6

From: PowerNovo: de novo peptide sequencing via tandem mass spectrometry using an ensemble of transformer and BERT models

Figure 6

(1) Performance of DeepMSPeptide, PepFormer and Peptide BERT block (PowerNovo) for peptide detectability prediction with the test dataset from the GPMDB database (Homo sapiens (20 K) + decoys (10 K)). (2) T-SNE visualization of the latent embeddings space for Peptide BERT block. The visualization results indicate that our model can efficiently capture high-latent discriminative information, improving the predictive performance. Plot was obtained using python 3.9 script and matplotlib 3.8.0 library.

Back to article page