Tracers and tagged molecules¶

Tracer experiments — stable isotopes, radiolabels, photoactivatable and photobleachable fluorophores — are among the most informative things you can do to a metabolic or trafficking system, because they measure fluxes rather than pools.

ProcessDB supports them directly. You do not build a second model for the label: you say which molecule is tagged, describe how the label was introduced, and ProcessDB derives the tracer equations alongside the chemical ones.

The tracer formulation¶

For every traced process, the tracer flux is the total flux times the label fraction of the donor pool:

P_tr  =  P · S_tr / (S + ε)

That is the classical formulation: the tracer system is linear in the tracer quantities even when the chemical model is nonlinear, and labelled and unlabelled material always share the same rate laws — which is the assumption tracer experiments rest on anyway.

Creating a tagged molecule¶

In the Molecules and Complexes browser, select a simple molecule and press Create Tagged Molecule (or use the right-click menu). Name it for the label — [U-13C]glucose, GFP-albumin.

A tagged molecule never appears as a state in the diagram. It exists so that experiments and MOEs can talk about the labelled form of a molecule that is already in the model.

Tracer protocols¶

On an experiment's Tracer Protocols tab, the left list holds the tagged molecules this experiment uses; press Add to include one. The right side then offers a tab per kind of tracer input, each identifying the tagged molecule, the target state, and its place:

Photobleaching and photoactivation generate the extra machinery those experiments need — bleached and retained fractions, a FRAP readout, a normalisation constant, an activated pool with its own equations — so a FRAP or photoactivation time course can be fitted directly.

Choosing what carries the label¶

Not every part of a network carries every label. In the Tracers tab of a Model of Experiment, choose a Tagged molecule and two checklists appear beside each other: Traced States and Traced Processes. Checked means traced.

The candidates are computed by following the label forward: starting from the states containing the tagged molecule, along reactant → process → product edges, and onward. Modifier roles are never followed — an enzyme cannot gain or lose label through a reaction it catalyses — so enzymes, activators, and inhibitors do not appear.

Everything reachable starts checked. Unchecking is how you route the label:

• a traced process takes its label fraction from its traced reactant and passes label to its traced products;
• unchecking a state stops the label there;
• unchecking a process removes its tracer flux equation entirely.

Rows the label cannot reach from this tagged molecule are listed as well, marked [not reachable from this tracer], as are rows left from an earlier version of the model, marked [not in the model]; either can be cleared. Your choices are saved with the MOE by the ordinary Commit and always win over the automatic defaults, so a routing you worked out once is never quietly recomputed.

The tab needs the model's structure. Opened on its own, a MOE has not loaded it, and the tab says so — open the Model Realization. If none of the MOE's experiments has a tagged molecule yet, the tab says where to add one: an experiment's Tracer Protocols tab.

Reading the results¶

Traced states and processes appear in the simulation output as their own variables and can be plotted and fitted like any other. Measurements of labelled quantities are associated with those variables, exactly as for unlabelled ones — so a labelled-fraction time course from a mass spectrometer constrains the model in the same machinery as everything else.