mzio GmbH announces the launch of mzassist, a vendor-neutral AI and automation control layer for multimodal mass spectrometry data processing on its trusted, open-source mzmine platform.
Image Credit: mzio GmbH
mzassist uses the Model Context Protocol (MCP) open standard introduced by Anthropic in November 2024, now supported by many AI agents and platforms, to connect mzio’s established, vendor-agnostic mzmine platform to a user’s approved MCP-compatible AI assistant. For the lab user, this simplifies set up and implementation of complete mass spectrometry data processing workflows in natural language; importing raw data, selecting templates, configuring parameters, running batches and summarizing results to increase operational consistency and free up lab time to focus on scientific input.
mzio shared exclusive preview access to mzassist with the scientific community in a workshop at the 8th Munich Metabolomics Meeting on October 13th, 2026, ahead of public release, giving researchers hands-on experience with the technology in LC-MS, GC-MS and MALDI-MS and gaining valuable feedback.
The mzassist approach draws a clear line on security and scientific control in the laboratory: the raw mass spec data stays on the lab’s own computer, and users control the information that is shared with the AI via opt-in / opt-out protocols. Because the AI tool is only active when enabled by the user, the user decides at every stage how much to hand over, either delegating the entire workflow or enabling manual intervention at any stage. Parameter decisions are made by deterministic mzmine algorithms rather than the language model, and labs connect the AI assistant that their IT and data protection teams have already approved, increasing accessibility and speed for mass spectrometry data management.
“The AI story in lab informatics has now moved from ‘can we use it?’ to ‘how do we use it responsibly?’” Dr. Ansgar Korf, CEO and co-founder of mzio comments. “mzassist brings together two key factors in mass spectrometry data management: the reliability of our long established, trusted mzmine system to handle the scientific decisions, and new accessibility through AI. The lab chooses which AI assistant to trust, and mzmine stays the authoritative environment for the data analysis itself. That combination makes AI practical in a mass spec workflow without compromising reproducibility or the vendor-neutral position that mzmine has held since 2005.”
mzassist can also help lab users get started faster with both existing and new analytical methods, explaining processing steps, algorithms and parameters, and supporting spectral library building. The parameter estimator analyzes a selection of the user’s raw data files and derives suitable processing parameters, and the auto optimizer then benchmarks and identifies the settings that perform best on the actual data.
“Parameter estimation and optimization in mzassist are data-driven algorithms, not guesswork by a language model,” adds Dr. Steffen Heuckeroth, CTO and co-founder of mzio. “The estimator derives parameters directly from the signals in a lab’s raw data, with consistent results every time. Users and AI agents can both call on these tools to get the same measurable basis for deciding which parameters to use, and every analysis started through mzassist is documented as a complete mzmine workflow file that records each processing step and parameter, meaning AI assistance increases productivity while keeping results reproducible and traceable."
mzassist will be available with mzmine 4.11 October 2026. Laboratories can connect the AI tool that fits their own IT, data protection and compliance requirements. It is included in mzmine COMMUNITY for academic users and in mzmine PRO for commercial customers, supporting mzio’s mission to empower researchers and analysts around the world with reliable tools for mass spectrometry data analysis.