Computational models

Most of our research is based on modeling biochemical networks. This page lists most of the models developed in the group throughout the years.

Most of our models are based on COPASI and Gepasi. The models provided below can be run in your browser using ShinyCOPASI, that is a smaller web-based version of COPASI. From there you can also download the model files to your computer and run them directly with the full COPASI.

Iron Mouse Pv4

Run this model

This is the fourth iteration of our mouse whole-body iron model. It includes loss of iron through the bile (from liver compartment), as suggested by the data published in Mercadante et al. (2018) Gastrointestinal iron excretion and reversal of iron excess in a mouse model of inherited iron excess. Haematologica 104:678-689.

Iron Mouse Pv3

This is the third iteration of our mouse whole-body iron model. It describes the flow of iron between organs in the body and was validated against a number of diseases and interventions (below are a couple of these). Published in Parmar JH, Mendes P (2019) A computational model to understand mouse iron physiology and diseases. PLoS Computational Biology 15(1):e1006680.

Noise in gene expression

Run this model

This is model of gene expression including reversible formation of an inhibitory secondary structure in the mRNA 5'UTR, showing that noise at this level can be as significant as promoter-driven noise. Published in Dacheux et al. (2017) Translation initiation events on structured eukaryotic mRNAs generate gene expression noise Nucleic Acids Research 45:6981-6992.. (This model is run with a stochastic simulation algorithm).

Yeast glycolysis pathway

Run this model

While many models of yeast glycolysis exist, in 2013 we published a model based on a consistent kinetic characterization of all its enzymes. We measured the kinetic parameters of all enzymes in similar conditions, estimated the enzyme concentrations by quantitative proteomics (QConCat), and built a sequence of models through cycles of experiment and modeling. The final model (18th iteration) is available here. This was published in Smallbone et al. (2013) FEBS Letters 587:2832-41.

Yeast glycolysis and pentose phosphate pathway

Run this model

Also in 2013, we expanded the yeast glycolysis model above with the pentose phosphate pathway, which we also characterized all its enzymes kinetically in vitro, while estimating their concentrations by quantitative proteomics. The paper was published as a manuscript and even though it received one round of reviews, one of the requests made by the reviewers was no longer logistically possible to address. The paper remains as a non-peer reviewed preprint, but has since received a number of citations. The COPASI model is available here or can be run directly on your browser too.

HIV Protease inhibitor

A HIV protease inhibitor model which is an example parameter estimation problem originally published in Kuzmic (1996) Anal. Biochem. 237:260-273, and used by us in Mendes & Kell (1998) Bioinformatics 10:869-883.

Model file: HIVProt.cps and associated fluorescence data file: HIVProt.txt.