Home     Screenshots     Downloads     Help Manual     Contact

 

SimCell Help

1. Introduction
2. Create
    2.1 Objects
  2.1.1 Membrane
  2.1.2 DNA
  2.1.3 Gene
  2.1.4 Protein
  2.1.5 Membrane Protein
  2.1.6 Small Molecule
    2.2 Interactions
  2.2.1 Transport
  2.2.2 Bind and Stick
  2.2.3 Touch and Go
    2.3 Others
3. Menu
  2.1.1 File
  2.1.2 Edit
  2.1.3 Run
  2.1.4 Help
4. Simulator
    4.1 Screenshot
    4.2 Graph
    4.3 Region Graphing

 

Introduction   

            In building our DCA cell simulator – called SimCell, we tried to create a very general structure that would allow the accurate simulation of most cellular pheonomena (transport, enzyme kinetics, metabolism, reaction-diffusion, signaling, genetic circuits and transcription/translation).  This required selecting a set of molecular components and choosing a collection of interaction rules that would be sufficiently diverse to describe the events that typically go on both inside and outside a cell.  Based on these considerations we identified 4 kinds of component molecules: 1) small molecules (metabolites, ligands), 2) membrane proteins (which can only exist in membranes), 3) soluble protein/RNA molecules and, 4) DNA molecules (which are non-mobile).  Because cells are composed of subcellular organelles and because some simulations require compartmentalization of reactants, products or metabolites, we also introduced a fifth type of molecule or super-molecule – the membrane.  In our model, membranes are non-mobile entities that describe boxes or borders that may be permeable or impermeable to certain molecules.  These membrane super-molecules may be used to define cytoplasmic compartments, periplasmic compartments, nuclear compartments or other subcellular organelles.

In SimCell the motion for all objects (except membranes and DNA molecules) is performed in a step-wise fashion over the space defined by a regular 2-dimensional grid.   Within the grid each object or molecule occupies a square, typically measuring 3 nm on a side.  The choice of 3 nm as the grid increment is based on a number of criteria.  For instance, the diameter for an average protein is approximately 3 nm, the average width of a cellular membrane is approximately 3 nm, the average width of a supercoiled DNA molecule is about 3 nm and the average velocity of a macromolecule in a cell is approximately 3 nm per millisecond [Sundararaj et al., 2003; http://redpoll.pharmacy.ualberta.ca/CCDB cgi-bin/STAT_NEW.cgi].  This macromolecular diffusion rate also helps to define a time step that is appropriate for these mesoscale simulations – namely a millisecond.  Macromolecules such as proteins or RNA molecules are allowed to occupy only a single square at a time, whereas up to 100 small molecules (which typically measure 0.3 nm across) are permitted to occupy a single square.  This multiple occupancy rule for small molecules avoids the need to reduce the grid size and time steps by a factor of 10. This simplification also makes the simulations run much faster.  Small molecules are known to diffuse at approximately 50 nm/ms [Jacobson and Wojcieszyn, 1984], or approximately 10 times faster than macromolecules.  To accommodate their faster diffusion rate, SimCell performs 10 small molecule movements/interaction checks for each macromolecular movement step.  To provide additional flexibility, SimCell users are free to define the grid size and geometry, depending on the type, number and complexity of the reactions, models or components.   The default geometry and grid size is a square composed of 100 x 100 elements.

For mobile molecules such as protein/RNA molecules, small molecules and membrane proteins, the initial spatial location (uniformly distributed, North, East, West, South bias, outside, inside cell or nucleus), number of molecules, molecule colour (for viewing), molecule properties (permeability with respect to the membrane, velocity, complexation state, decay rate, creation rate, etc.) and molecule label are all defined by the user during the initial set-up phase.  These properties, which are entered into pop-up property cards, may be changed at any time during a simulation run.  It is important to note that mobile molecules also have an optional “creation” rate (linear or exponential) that can be entered into their property card (the default value is always 0).  The creation option may be used to describe a general or external source for nutrients, metabolites or macromolecules – without the need to fully map out superfluous metabolic or transcriptional pathways.  The molecular decay rate (linear or exponential) may be used in a similar manner to describe external or internal sinks needed to eliminate individual or monomeric molecules.  However, the decay rate can also be used to model the disassociation of complexes composed of two or more molecules

Static or non-mobile entities in SimCell are treated slightly differently than mobile molecules.  For genes, operons or gene/promoter elements residing on the cell’s DNA, users must designate the location of these components (via dragging and dropping a coloured gene icon onto the DNA strand) once the DNA molecule is drawn.  Properties for the specific genes may also be described, including the transcription rate the number of genes, the gene order and the type of protein/RNA molecules that may be transcribed from the selected gene or operon. Protein transcription, in the current version of SimCell, skips the steps of RNA production, processing and ribosomal binding.  This abstraction is done to both simplify the model building process and accelerate the simulation run.  

To facilitate interactive model design SimCell uses its own graphical modelling tool, written in Java.  This model rendering tool allows a schematic description (using single representative molecules or entities) of the model components to be drawn on a simple canvas using a set of drag-and-drop palette tools (see Figure 1a).  As each molecular object is moved onto the canvas, the user is prompted with a pop-up window or dialog box to fill in specific information about that molecule or entity.  The molecular objects and their property cards can be edited or deleted from the canvas at any time during the model construction process.  Additionally, the SimCell GUI permits a variety of interaction arrows or connectors to be drawn between molecular entities.  These interaction arrows are used to describe the allowed pairwise interactions and their interaction consequences.  Through SimCell’s model rendering tool, new models may be generated or old models may be uploaded, edited and saved. The program also allows general notes or comments to be added about each saved model.  Prior to generating a model through SimCell’s model rendering tool, it is suggested that users compile a list of all enzymes, proteins, ligands, reactants and products needed to describe the model.  A common error among first-time SimCell users is the tendency to neglect to include products or resultant complexes in their schematic diagrams.  Once the model’s schematic wiring diagram is completed and the necessary fields filled in, the simulation is ready to be run. 

To initiate the run phase, SimCell maps the schematic pathway or wiring diagram onto its simulation grid.  This involves generating the 2D grid, defining the compartments, placing the requisite number of molecules over the grid and assigning initial velocities/directions to each molecule.  Error or consistency checks are performed during this mapping phase to ensure that all objects have a place to go, that key information is not missing and that the interactions and the interaction grid are logically consistent.  If any errors or inconsistencies are found, the user is prompted to edit the offending value in the pop-up property or interaction windows.  Once the model mapping is complete, a simulation viewer is launched in a separate window.