Skip to content

Fluxion v5.5.2

Choose a tag to compare

View all tags
@nicgustafson1 nicgustafson1 released this 24 Jul 01:12
v5.5.2
a90fc59
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Fluxion 5.5.2

What's New

  • Glycogen system added to the model. This system is responsible for glycogen synthesis and breakdown. It is a key component of energy storage and release in cells.
  • The malate-aspartate shuttle is now modelled properly, as the six reactions it actually is, instead of a single placeholder. Its aspartate/glutamate carrier moves net charge, which is what makes the shuttle run one way in a respiring cell.
  • Compartments now have realistic volumes. The intermembrane space is a thin shell rather than a space the size of the cytosol, so material moved into it no longer disappears from the cell's working pools.
  • The mitochondrial membrane potential is now a real voltage, shown in millivolts, calculated from charge separation across the membrane. Every reaction that moves charge responds to it.

Improvements

  • Re-evaluation of literature values for many reactions. Largest fix being malate dehydrogenase.
  • Reversible reactions now hold their equilibrium constants at every concentration. Previously they only did so while metabolites stayed well below their Km values, and drifted badly above that: triose phosphate isomerase sat 96x away from its correct equilibrium, which trapped most of glycolysis' carbon in DHAP and stalled the pathway.
  • The electron transport chain has been rebuilt on published midpoint potentials, with the correct proton and charge stoichiometry for each complex.
  • ATP synthase is a single reaction driven by the protonmotive force, replacing a two-step rotor-and-catalysis pair joined by an invented intermediate.
  • The adenine nucleotide translocase, phosphate carrier and pyruvate carrier now respond to the membrane potential and pH gradient instead of having those effects written in as fixed constants.
  • Lactate leaves the cell with its proton, as it does through the real transporter. Without this the cell acidified without limit under load.
  • Phosphofructokinase is now inhibited by falling pH, the feedback that stops glycolysis outrunning its own downstream capacity during hard work.
  • Enolase and adenylate kinase equilibrium constants corrected. Enolase was carrying fumarase's value.
  • Resting concentrations updated to measured muscle values, including ATP, free ADP and AMP, and the citric acid cycle intermediates, which previously all started at zero.
  • Guided exercise workloads rebalanced so the aerobic, hypoxic and anaerobic conditions separate clearly across phosphocreatine, lactate, membrane potential and ATP.
  • Simulations run roughly twice as fast, after switching to a solver better suited to the model. The change was made to fix the crash noted below, and the speedup came with it.

Bug Fixes

  • Reduced lag spike when using timed velocity changes.
  • Creatine kinase and pyruvate kinase could consume protons the cell did not have, because neither slowed down as free protons ran out.
  • Several reactions that produce NADH were not producing the proton that comes with it, and ATP synthase was not consuming the one that ATP synthesis takes up. Both left the model quietly creating or destroying acid.
  • The lactate transporter's reverse velocity was ten thousand times too high, which drove blood lactate slightly negative early in a run.
  • Restoring oxygen to a cell that had been running anaerobically could stop the simulation partway through, either with a "math domain error" or with the solver giving up. Recovery from anaerobic conditions is a demanding moment to calculate — the electron transport chain restarts against a collapsed membrane potential and everything moves at once — and both the arithmetic and the solver now handle it.

Known Issues

  • The membrane potential reads high, around 200 mV against a textbook 150-180. The total driving force across the membrane is correct, but the model cannot yet split it properly between the electrical and pH components, so too much of it shows up as voltage. Mitochondrial and cytosolic NAD+/NADH ratios are lower than measured values for the same reason.

SHA-256 checksum

d069afbe153c73162b501b477b121e0722647b03af9ab9822575e3b985aad1cb

Assets 5
Loading