Cracking the mass spectrometry code: Artificial intelligence for de novo peptide sequencing

AI is revolutionizing de novo peptide sequencing, enabling deep learning models to decode MS/MS spectra into amino acid sequences—without relying on reference proteomes. I'll introduce Casanovo, a transformer-based model that delivers state-of-the-art performance, accuracy, and generalizability across a range of proteomics applications. To accelerate innovation, I'll present a community benchmarking effort that reveals where current tools excel—and where they still fall short—empowering both developers and practitioners to push the field forward.