Virtual Cell at UConn


BioNetGen stand-alone


Sample models

 

  1. START WITH: Modeling domain-specific interactions and complex formation. simple.bngl model is a test model that is commented through and contain all latest features implemented in 2.0.48.
  2. START WITH: Modeling domain-specific interactions and complex formation: toy model 1. toy1.bngl model is a test model that is commented through. It consists of a monovalent extracellular ligand, a monovalent cell-surface receptor kinase, and a monovalent cytosolic adapter protein. The receptor dimerizes through a receptor-receptor interaction that depends on ligand binding. When two receptors are juxtaposed through dimerization one of the receptor kinases can transphosphorylate the second receptor kinase. Apapter protein A can bind to phosphorylated receptor tyrosine. Input file specifies 3 initial species and 5 rules. Generated reaction network consists of 14 species (all complexes such a dimer with two adapter proteins bound, a dimer with just one tyrosine phosphorylated, etc). The model reports totals such as total amount of phosphotyrosines and total amount of A bound to phosphotyrosines.
  3. START WITH: Modeling polymers: interaction of bivalent receptors with bivalent ligands. Formation of chains and rings of a finite length. bivlig_bivrec.bngl model describes chains and rings formed by interactions of bivalent ligands and bivalent receptors (picture). There are three rules: ligand binding to a receptor, chain elongation and ring closure. Model generated by BioNetGen includes all chains and rings consisting of no more than 10 ligand and 10 receptors - a total of 40 species and 360 interactions among them are generated. Warning - the model may run up to 5 minutes, depending on the Virtual Cell load.
  4. Modeling a reaction cheme. kholodenko_1999.bngl model follows a paper by Kholodenko et al. (1999) and describes a reaction scheme among explicitely specified species (only species and reactions related to Sos recruitment are included). All species (18) and interactions (37) are explicitely specified in BNGL language, as well as simulation parameters. Simulation provides timecourses for all 18 species as well as totals corresponding to experiements, such as total phosphorylated EGFR, total phosphorylated Shc, etc (eq. 2-7 and eq.13-17 of Kholodenko et al.). We call these totals "observables". SBML file is generated and can be imported into your favorite simulator.
  5. Modeling domain-specific interactions and complex formation: toy model 2. toy2.bngl model is a test model that is a modification of toy1.bngl model. Receptors do not have a kinase domain, but rather a cytosolic kinase transphosphorylates receptor tyrosines. The receptor dimerizes through a receptor-receptor interaction that depends on ligand binding. The adapter binds the receptor and the kinase. When two kinases are juxtaposed through binding to receptor-associated adapter proteins, one of the kinases can transphosphorylate the second kinase. Due to considering of different dimeric complexes, the total number of species and reactions generated by BioNetGen is 25 and 101 respectively. However, input file specifies only 5 initial species and 8 rules.
  6. Modeling domain-specific interactions and complex formation: early events in EGF receptor signaling. blinov_2006.bngl model follows a paper by Blinov et al. (2006) and describes essentially the same interactions as in the model by Kholodenko et al., but considering independent binding sites of interacting proteins and accounting for all potential complexes and phosphoforms that can be generated during a response to a signal. This model predicts the dynamics of 356 molecular species, which are connected through 3749 unidirectional reactions. These species and reactions are not specified explicitely, but generated from the set of rules specified in blinov_2006.bngl file in BNGL language. New predictions include temporal patterns of autophosphorylation for different tyrosine residues of EGFR.
  7. Modeling genetic oscillator. vilar_2002.bngl model follows a paper by Vilar et al.(PNAS, 2002). The model is done under assumption that DA and DR are parts of the same DNA strand. Both genes are defined as components of a single multi-component DNA D(p,r), where p is a activator A promoter region and r is a repressor R promoter. Degradation is just R-> trash, trash is a molecule type. A can bind to either activator promoter or repressor promoter, but not to both. There are two more extensions of this model done in BioNetGen language. vilar_2002b.bngl model describes a mechanism where A can bind to activator promoter and repressor promoter independently. vilar_2002c.bngl model extends to 2 activator promoters and 1 repressor promoter, and A can bind independently to any of them.
  8. Modeling chemistry: the conversion of 7,8-dihydrofolate (A) and NADPH (B) into 5,6,7,8-tetrahydrofolate (C) and NADP+ (D), catalyzed by the enzyme dihydrofolate reductase (E). chemistry.bngl model represents the overall reaction A+ B = C + D, which is a culmination of the 13 elementary reaction steps comprising 13 chemical species (Fishtik et al., , Ind. Eng. Chem. Res.; 2006; DOI: 10.1021/ie050814u). The reaction graph represented in the (figure) can be represented as a set of 7 rules. Please note that this is just meant to be an illustrative example, and I'm not a practicing chemist. Therefore, please let me know about any mistakes.

More models, with fewer/no comments. Please contact blinov@uchc.edu for more details.


  1. MAPK cascade in yeast - no dimerization of Ste5. mapk-monomers.bngl
  2. MAPK cascade in yeast - dimerization of Ste5. mapk-dimers.bngl
  3. Notch pathway. notch.bngl
  4. Wnt pathway. wnt.bngl