CasanovoGUI puts deep-learning de novo peptide sequencing in a point-and-click desktop app

Researchers at the University of Washington, the University of Michigan and the University of Antwerp have released CasanovoGUI, a free desktop application that lets laboratory scientists run Casanovo, a deep-learning model for de novo peptide sequencing, without using the command line. The paper, "CasanovoGUI: A Cross-Platform Desktop Application for Deep Learning-Based De Novo Peptide Sequencing with Casanovo," was published in the Journal of Proteome Research on October 19, 2026. [1]

What de novo sequencing does

In a typical proteomics experiment, peptides are fragmented in a mass spectrometer, and each tandem mass (MS/MS) spectrum is matched against a database of known protein sequences. That approach cannot detect peptides that are not in the database. De novo sequencing reads the amino acid sequence directly from the spectrum instead, which the Casanovo developers describe as valuable for antibody sequencing, immunopeptidomics and metaproteomics. [2]

Casanovo, from the laboratory of William Stafford Noble, approaches the problem like translation: a transformer neural network converts the sequence of peaks in a spectrum into a sequence of amino acids. In a 2024 Nature Communications paper, the team reported training the model on 30 million labeled spectra and found that it outperformed several state-of-the-art methods on a cross-species benchmark. [2]

The usability gap

According to the new paper, Casanovo is distributed as a Python command-line program. Installing it, configuring the GPU and software dependencies, and setting parameters by hand can be difficult for many bench scientists and are a recurring source of errors. Interpreting and validating the resulting predictions is a further hurdle. [1]

What CasanovoGUI adds

CasanovoGUI is an open-source, Java-based application that exposes Casanovo's main analysis functions through a point-and-click interface on Windows, macOS and Linux. On first use, it installs a private Python environment and a GPU-matched build of Casanovo automatically, so no prior software setup is needed. [1] The project's documentation says the installers bundle their own Java runtime. [3]

The interface gives access to Casanovo's configuration parameters and streams live progress. It also builds result checking into the same window: annotated spectra with per-residue confidence scores are shown in the PDV viewer, and de novo peptides can be mapped back to a reference proteome with mismatch tolerance. [1] PDV is an existing open-source proteomics data viewer first described in Bioinformatics in 2018. [4]

The project page lists separate tabs for de novo sequencing, database search, evaluation and model training, plus a View tab for exploring results. [3] The software is released under the GNU General Public License v3.0, and version 1.4.0 was published on GitHub on September 12, 2026. [3] A step-by-step guide is part of the Casanovo documentation. [5]

Why it matters for analytical work

De novo sequencing is one way to check what a spectrum actually contains rather than only what a database expects. Per-residue confidence scores and proteome mapping give analysts a direct view of how well each predicted sequence is supported. For laboratories that rely on mass spectrometry to confirm the identity of synthetic peptides, tools that make this kind of check easier to run and easier to review are a practical step forward.

Related: De novo sequencing candidates re-ranked by spectrum similarity · A billion-scale mass spectrometry corpus · Basic terms · Reading a COA · Peptide Academy

All BioLabs Research products are sold strictly for laboratory research use only. Not for human or veterinary use. This article summarizes published research and makes no claims about any BioLabs product.

For research use only. Not for human or veterinary use. Not a drug, food, or cosmetic.