Welcome to Maud’s documentation!#

Maud is a tool for fitting Bayesian statistical models of metabolic networks.

Maud’s distinguishing features include:

  • Scientifically accurate representation of phenomena including enzyme kinetics, allosteric regulation, competitive inhibition, phosphorylation, knockouts and transported charges.

  • Guaranteed consistency with thermodynamic and steady state constraints.

  • A statistical model allowing inference consistent with both information from experiments and pre-experimental background information.

  • Prediction of steady state concentrations and fluxes given out-of-sample boundary conditions.

More practically speaking, Maud is a Python application that you can use from the command line. Maud provides the command maud sample, which loads data from an input folder containing toml files representing a metabolic model, experimental measurements and background information, validates this input and passes it to a statistical model specified in using the probabilistic programming language Stan. When the statistical computation is finished, Maud converts the results into a convenient format and saves them in an output folder.

To find out how to install and use Maud, check out the Getting started page.

For a guide to creating inputs for Maud, see the Specifying input data page.

For what to do when things inevitably go wrong, read the Troubleshooting page.

For detailed discussion of the scientific assumptions implicit in Maud’s statisical model, there is the Theory page.

For reference information about Maud’s command line interface, see the CLI reference page.

To find out how to contribute to Maud, read the Contributing to Maud page.