psamm.gapfill – GapFind/GapFill

Identify blocked metabolites and possible reconstructions.

This implements a variant of the algorithms described in [Kumar07].

exception psamm.gapfill.GapFillError

Indicates an error while running GapFind/GapFill

psamm.gapfill.gapfind(model, solver, epsilon=0.001, v_max=1000, implicit_sinks=True)

Identify compounds in the model that cannot be produced.

Yields all compounds that cannot be produced. This method assumes implicit sinks for all compounds in the model so the only factor that influences whether a compound can be produced is the presence of the compounds needed to produce it.

Epsilon indicates the threshold amount of reaction flux for the products to be considered non-blocked. V_max indicates the maximum flux.

This method is implemented as a MILP-program. Therefore it may not be efficient for larger models.

Parameters:
  • modelMetabolicModel containing core reactions and reactions that can be added for gap-filling.
  • solver – MILP solver instance.
  • epsilon – Threshold amount of a compound produced for it to not be considered blocked.
  • v_max – Maximum flux.
  • implicit_sinks – Whether implicit sinks for all compounds are included when gap-filling (traditional GapFill uses implicit sinks).
psamm.gapfill.gapfill(model, core, blocked, exclude, solver, epsilon=0.001, v_max=1000, weights={}, implicit_sinks=True, allow_bounds_expansion=False)

Find a set of reactions to add such that no compounds are blocked.

Returns two iterators: first an iterator of reactions not in core, that were added to resolve the model. Second, an iterator of reactions in core that had flux bounds expanded (i.e. irreversible reactions become reversible). Similarly to GapFind, this method assumes, by default, implicit sinks for all compounds in the model so the only factor that influences whether a compound can be produced is the presence of the compounds needed to produce it. This means that the resulting model will not necessarily be flux consistent.

This method is implemented as a MILP-program. Therefore it may not be efficient for larger models.

Parameters:
  • modelMetabolicModel containing core reactions and reactions that can be added for gap-filling.
  • core – The set of core (already present) reactions in the model.
  • blocked – The compounds to unblock.
  • exclude – Set of reactions in core to be excluded from gap-filling (e.g. biomass reaction).
  • solver – MILP solver instance.
  • epsilon – Threshold amount of a compound produced for it to not be considered blocked.
  • v_max – Maximum flux.
  • weights – Dictionary of weights for reactions. Weight is the penalty score for adding the reaction (non-core reactions) or expanding the flux bounds (all reactions).
  • implicit_sinks – Whether implicit sinks for all compounds are included when gap-filling (traditional GapFill uses implicit sinks).
  • allow_bounds_expansion – Allow flux bounds to be expanded at the cost of a penalty which can be specified using weights (traditional GapFill does not allow this). This includes turning irreversible reactions reversible.