It is certainly true that the innate complexity of Biological systems often exceeds the cognitive "purchase" of the human mind. For example, it is one thing to consider the growth kinetics of a single living cell and quite another to begin to empirically model the biochemical processes associated with biological systems at the cellular level. This is an essential difference. The theoretical and algorithmic models under development are spatial and temporal rather than purely biochemical. These models permit a virtual cell to cycle independently within a "Digital-BioDomain", ultimately forming a complex system of interacting cell populations.
The cycle kinetics for this project will be rendered in 3D and 4D space (within the Domain), allowing variable and compressible time intervals (Quadri-dimmensional constructs). A simultaneous display of cell cycle information will accompany the graphical display of the tumor growth. Ultimately the ability to link digitally scanned actual tissue tumors stained quantitatively for DNA and create virtual "objects" or containers within which the virtual cancer cells could grow would permit correlation for Macroscopic and Histologic features with the kinetic growth models in an unprecedented fashion. The opportunity to render these combined growth models as actual objects in real space gives an opportunity to study cancer growth from entirely novel perspectives.
It is clearly an equal challenge to permit many cells to cycle interactively yet independently.
Added to that challenge is the task of defining entirely new populations which "interact" with each other dynamically for example tumor cells and virtual lymphocytes. Fortunately, within this model each of these functions can be considered an "extension" of the same set of algorithms.
Although each cell has its own unique properties and attributes they can all follow the same "Rules to live by" within the Digital-BioDomain. It should also be recognized that the same algorithms may be applied to model molecular interactions as well as entire populations dynamics.
Creating the Digital-Domain would provide an environment in which to conduct various aspects of biological (and cancer) research in a heretofore-unprecedented manner. It would also serve to provide algorithms, which are capable, by extension of class description, of evolving into more complex models over time.
These positive and negative effects could "Invert" under given
conditions or under other conditions "Sum-up" to trigger an event in another set of "Modifiers". For example a Digital-Receptor on the cytoplasm is seen by the program as a positive or negative value that can be summed, graded or otherwise integrated into the overall cell cycle and in turn effect a cell modifier for a specific nuclear attribute (or vice versa).
Using this approach the complex interaction of cell markers can be modeled in a relatively straightforward manner. In effect reducing the complex interaction between cells and multiple biological markers to the summation of positive and negative values for the given cells.
It then becomes possible to simulate the cumulative and interactive effects of Digital biological-modifiers (mitogens and antimitogens), dynamically and interactively, on digital cell cycles and virtual tissue population kinetics.
If it is possible to manipulate both mitogenic and antimitogenic modifiers with a population of virtual cancer cells in a Digital-BioDomain then it would then be possible to evaluate the interaction of "Digital Chemotherapeutic" substances which have known cell cycle properties. For example the difference between Cell Cycle specific and Cell cycle non-specific Chemotherapeutic substances would be represented as a positive versus a negative "modifier" for the entire population of cells throughout all phases of the cell cycles or specific cell cycle phase such a S-Phase of Late G1/G0 components.
To illustrate this point consider a substance that is know to prevent
cells from leaving late G1 phase of the cell cycle for example Tamoxiphen.
This effect would be represented by a variable governing the probability
of a cell to leave the late G1 phase of the cell cycle. If the probability is relatively low then
the specific cell is less likely to leave the late G1 phase of the cycle.
If the probability is high then the cell is more likely to leave the late G1 phase of the cell cycle. Interestingly it becomes clear that some cells will "slip though" based on probability alone.
Different variables would represent different Digital Chemotherapeutic substances. The interaction of these substances would then be a dynamic function of the model. This approach presents a unique opportunity to vary the amount and timing of these various combinations of virtual chemotherapeutic substances to determine the overall impact on the cell cycle kinetics in real time. In fact it would then be possible to "run" multiple and simultaneous variations of time sequence studies with variable (progressive/regressive) dose-regimen combinations.
It should be remembered that normal tissue would be modeled simultaneously to determine the potential impact on non-neoplastic components.
In exploring the concept of Digital-BioDomains several potential applications became apparent. One of the most interesting was the possibility of providing empirical
information which could be used in optimizing pharmaceutical clinical trials. This information would be available prior to implementing the protocol. It would be possible to run multiple combinations of cell cycle specific and non-specific Digital Chemotherapeutic substances across many variations of neoplastic growth kinetics. This may provide information helpful in determining the best possible combination of Bio-Physiologic factors for a proposed research population.
For example, in breast carcinoma, it might be possible to evaluate the optimal combination of estrogen/progesterone receptor (quantitatively), gene profiles, and S-Phase fraction needed to optimize the effect of a given chemotherapeutic substance. This might be accomplished prior to initiation specific clinical trial.
In this application it seems reasonable that a cell cycle model which approximates actual biological processes could provide empirical data to be used in configuring and prospectively optimizing pharmaceutical clinical trials prior to involving human subjects.
Model specific aspects of neoplastic growth: The digital carcinoma will be rendered in three dimensional space using fully rotational graphics. Cells in different phases of the cell cycle will be represented as colored spheres (not unlike a molecular model). Cells in G1/G0 may be represented as blue spheres, S-phase cells as Orange spheres; G2/M in green etc. Cells that are in transition, have died or those that have undergone a specific program related "change" can be flagged individually by the program.
Since these processes will be rendered in three dimensional space each cell would necessarily be assigned a point or range of points for it to occupy in that space. It would then be possible to define points in space where the neoplastic cells could not occupy. For example a set of points could define a cylinder. The cylinder would represent a duct, analogous to a Ductal-Lobular unit in human breast tissue. Within the duct it would be possible to "grow" an intraductal carcinoma. It would also be possible to redefine a subset of points in the duct wall where the neoplastic components could extend out into adjacent extra ductal space. This "Hole" in the wall could realistically model early invasive carcinoma. Cells in Extra-Ductal space could be restricted by different cell cycle parameters to remodel growth kinetics of invasive components.
A vascular component could be modeled as easily. Neoplastic cells would be more likely to grow in a direction of a digital vascular segment. In fact the neoplastic components could grow their own vascular supply in a relative approximation of Tumor angiogenesis. Remembering that this model exists within the computer it would even be possible to simulate thrombosis (or even tumor embolization) of an intra-tumoral vascular segment. The same algorithms could be extended to include metastasis and Immune responses.