SBML import: conservation laws for non-constant species #1669
Conversation
Also: Modified the run-python-tests script to work also outside of Python virtual environments (venvs) by installing pip packages to the user's home directory.
Also, debugging null_space(...) call for stoichiometric matrix.
| @@ -347,7 +347,8 @@ def sbml2amici(self, | |||
| sbml.L3P_PARSE_LOG_AS_LN | |||
| ) | |||
| self._process_sbml(constant_parameters) | |||
| if self.symbols.get(SymbolId.EVENT, False): | |||
| if self.symbols.get(SymbolId.EVENT, False) \ | |||
There was a problem hiding this comment.
what's the issue with heavisides in the flux vector, shouldn't break any conservation laws should it?
There was a problem hiding this comment.
Right. That "fix" was too specific for an SBML test case and misses the point.
The actual problem are conservation-law/x_rdata-dependent root functions:
AMICI/python/amici/ode_export.py
Lines 111 to 115 in 067826d
Should be addressed accordingly.
There was a problem hiding this comment.
Let me do that in a separate PR shortly.
python/amici/sbml_import.py
Outdated
| assert len(state_idxs) == len(coefficients) | ||
|
|
||
| # choose a state that is not already subject to removal | ||
| # TODO is this necessary or can we just take the first one? |
There was a problem hiding this comment.
with toposort + hierarchical expressions, I don't think this is necessary and we can just pick the first
There was a problem hiding this comment.
Among the states involved in that CL, we still need to choose a state that has not been eliminated yet. Confused myself.
python/amici/conserved_moieties.py
Outdated
| interactions of metabolites and reactions, and matrix of interaction | ||
| """ | ||
| dim = len(matched) | ||
| MIN = 1e-9 |
There was a problem hiding this comment.
my feeling is that MIN/MAX should be consistent across the different functions?
| T1 = initial_temperature | ||
| e = math.exp(-1 / T1) | ||
| while True: | ||
| en = int(random.uniform(0, 1) * dim) |
There was a problem hiding this comment.
I think it would be great if the determination of conservation laws was deterministic. Use a fixed random seed here?
There was a problem hiding this comment.
Yeah, I guess that makes sense.
Co-authored-by: Fabian Fröhlich <fabian_froehlich@hms.harvard.edu>
|
SonarCloud Quality Gate failed. |
Extends identification conservation laws during SBML import to non-constant species based on the algorithm proposed in [1].
Parameterized stoichiometric coefficients are not supported. Neither are models with events/Heaviside/piecewise or RateRules. Conservation laws for non-constant species are completely ignored in this case.
Experimental feature: Enable by setting environment variable
AMICI_EXPERIMENTAL_SBML_NONCONST_CLSto any value.[1] De Martino A, De Martino D, Mulet R, Pagnani A (2014) Identifying All Moiety Conservation Laws in Genome-Scale Metabolic Networks. PLoS ONE 9(7): e100750. https://doi.org/10.1371/journal.pone.0100750