New Software Release: "MxlPy – Bridging Mechanistic Modelling and Machine Learning"

We are excited to announce the release of MxlPy , a Python package developed in our lab to support mechanistic learning - the seamless integration of mechanistic models with machine learning techniques. MxlPy is designed for researchers who need interpretable , data-driven , and reproducible modelling tools. It enables construction, calibration, ensemble generation, surrogate modeling, uncertainty quantification (including Monte Carlo analysis), and visualization of dynamic systems governed by ODEs. Originally built to meet the demands of our three large collaborative projects: GoPMF, MibiNet and CEPLAS, involving complex biological systems and rich experimental data, MxlPy reflects our belief that machine learning should complement , not replace, mechanistic insight in life sciences. The software was developed by Marvin van Aalst , our dedicated research software developer , whose role underscores the critical importance of specialized software engineering in modern research. The project also benefited greatly from the contributions of Tim Nies , who led the structural analysis and stress-tested the software’s logic and usability, and Tobias Pfennig , whose work inspired and justified the implementation of native Monte Carlo functionality.

Download MxlPy on Github: https://github.com/Computational-Biology-Aachen/mxlpy

View Documentation and Examples: https://computational-biology-aachen.github.io/MxlPy/latest/

If you use MxlPy, please cite our preprint on bioarxiv