For the first time in five years, I found an Antimony round-tripping bug. This isn't a priority, I just wanted to file it here so we don't forget about it.
The following code:
model advanced()
# Create two compartments
compartment compA=1, compB=0.5 # B is half the volume of A
species A in compA, B in compB
# Use the label `J0` for the reaction
J0: A -> B; k*A
# C is defined by an assignment rule
species C
C := sin(2*time/3.14) # a sine wave
k = 0.1
A = 10
# Event: half-way through the simulation,
# add a bolus of A
at time>=5: A = A+10
end
Gets round-tripped as:
// Created by libAntimony v2.9.4
model *advanced()
// Compartments and Species:
compartment compA, compB;
species A in compA, B in compB, $C;
// Assignment Rules:
C := sin(2*time/3.14);
// Reactions:
J0: A -> B; k*A;
// Events:
_E0: at time >= 5: A = A + 10;
// Species initializations:
A = 10;
B = ;
// Compartment initializations:
compA = 1;
compB = 0.5;
// Variable initializations:
k = 0.1;
// Other declarations:
const compA, compB, k;
// SBO terms:
// - Compartment SBO Terms:
default_compartment.sboTerm = SBO:0000410;
end
C gets converted into a boundary species, and the simulations are different. Roadrunner plots the value of C in the first example, but not in the second.
Antimony 2.9.4, Win64, libSBML 5.15.0
The SBO term was added by Tellurium, can be deleted