Matrix Science at ASMS 2024
The ASMS 2024 conference was held in Anaheim, California, on 2-6 June 2024. It was great to see many of you at our booth as well as the annual Breakfast Meeting! We presented three talks at the meeting. If you were unable to attend, or want to review any of the material that was presented, the talks are summarised below and you can also download the full slides and speaker notes from the workshop page.
Preview of Mascot Server 2.9
The next version of Mascot Server is in the last stage of the development cycle, and we hope to reach a 2.9 beta version soon. Here are the highlights of the upcoming release.
More machine learning. Everyone says they are adding more ML, so what does it mean for Mascot? The next version of Mascot will integrate MS2Rescore, which is a “Modular and user-friendly platform for AI-assisted rescoring of peptide identifications” developed at the CompOmics lab at the University of Ghent. MS2Rescore is a front end to DeepLC for retention time prediction and MS2PIP for MS/MS spectrum prediction. We will add dropdown menus to the format controls in Protein Family Summary where you can select a suitable model for your LC and mass spec. Then, click a button and Mascot will run the predictions in the background and integrate them into the results report.
Faster error tolerant search. The Mascot Error Tolerant (ET) search is a two-pass search. In the second pass, Mascot identifies unsuspected cleavage and unsuspected modifications. Currently, when you run an ET search, Mascot iterates through over 2000 modifications in the Unimod database, which can take a long time. In the next version, you will be able to select a subset of modifications to reduce the search time. For example, select only N-glycosylation or only chemical derivatives. This makes the search faster as well as more precise.
Results file performance. For many years, Mascot has saved database search results in a text-based file with a .dat extension. In order to solve certain performance bottlenecks in Mascot Distiller as well as the interactive reports in Mascot Server, we are introducing a new file format for search results. The Mascot Search Results (MSR) format is an SQLite file that contains exactly the same search data, except structured as a relational database. You will be able to download results in both the old and new format.
‘Quality of life’ improvements, bug fixes. There will be some smaller ‘quality of life’ improvements, as they are sometimes called, as well as bug fixes. The full list will be available in the release notes.
Strong backwards compatibility. Mascot Server 2.9 continues the tradition of strong backwards compatibility. Our goal is that you can always update to the new version and your existing pipelines and workflows continue to work normally. The new version will be able to open results from any previous version, and all the client APIs (application programming interface) stay the same. If your pipeline requires results in the old results file format (.dat file), these are still available through the client API. Data can be exported in all the supported formats, including Mascot CSV, Mascot XML, mzIdentML and mzTab.
Custom reporting with Mascot Distiller
Mascot Distiller 2.8 ships with 14 reports for quantitation, including ANOVA, hierarchical and K-means clustering, PCA and volcano plots. This covers many use cases, but not everything. The reports are written in Python, and you can easily create your own. We covered the basic steps and how to get started. The user interface for reports is defined as an XML file, which Distiller adds to the reports menu, and the Python script is free to make any calculations, plot graphs, save files or call external processes. In addition to the slides, the full tutorial and example code are available in Tutorial: Creating custom reports in Mascot Distiller.
Automatically running reports with the Mascot Daemon Export Extender
Mascot Daemon includes a function that is sometimes overlooked: running an action automatically before or after a task. These are set up in the External Processes dialog. Daemon can run any command-line program after a task, but because the mechanism is so flexible, it comes with some complexity. We have written a wrapper called Mascot Daemon Export Extender (MDXE) that makes it very easy to automatically run Distiller reports at the end of a task. Simply uncomment the reports you need in the MDXE configuration file and add the program invocation to the External Processes dialog. MDXE is described in Mascot Daemon Export Extender as well as the slides.
Keywords: ASMS, error tolerant, machine learning, Mascot Daemon, Mascot Distiller