Virtual biochemistry

12 Feb 2010


The first step in creating the simulation was to define the metabolites. I limited myself to five metabolites, which I have given one-letter abbreviations. Carbon dioxide and nitrogen gas were assumed to be present in the environment at fixed concentrations. Oxygen was also present in the atmosphere at a fixed amount, but its concentration in cells could be changed by cellular reactions. The five metabolites in cells were:

  • C: fixed carbon, e.g. carbohydrates, such as sugars
  • N: fixed nitrogen, e.g. ammonium and amino acids
  • E: energy, e.g. ATP, NADH and NAPDH
  • O: oxygen gas
  • P: protein and other biomolecules


To keep the simulation simple (for once), the reactions I included in the simulation corresponded to pathways of multiple reactions in real life. Each of these reactions were catalysed by a single enzyme, which I named after the pathway itself or main enzyme in the pathway:

  1. Photosystem II: light → O + E
  2. Rubisco: 2E → C
  3. Catabolism: O + C → E
  4. Nitrogenase: 2E → N
  5. Anabolism: C + N → P

The reactions (numbered) are illustrated in this metabolic network.

Heterocyst chemical reactions

There were two further processes: diffusion and cell division. Diffusion caused the amounts of C, N and O to move between neighbouring cells, from where they were more concentrated to where they were less concentrated. Oxygen also diffused between cells and the environment, where its concentration was fixed. Cell division occurred when a cell accumulated 1000 units of P. At this point, 1000 units of P were removed and the amount of the other metabolites was halved. Then a daughter cell was created with the same amount of each metabolite as its mother.

Since the success of a genome was determined by the speed at which the filament reach a threshold number of cells, evolution was effectively selecting for cells (or filaments of cells) that maximised protein production.