Anyone that has ever tried to write a rate equation for an enzymatic reaction in a biological process knows that it often disperses into a series of overwhelming, unintelligible numbers. While singular biological pathways are hard enough to interpret with such conventional methods, whole cell systems that integrate several pathways become that much harder. Figures that result from physiological processes are complex enough to befuddle even the most experienced statisticians and powerful computers.
An innovative new approach to such studies is the use of “logic” as opposed to hard numbers based on the principles of computer programming.
In “executable biology,” as it’s being called, each part of a biological system corresponds to a piece of programming code. Hence the signals, interactions and thereby, the effects of the processes are explored using the corresponding computer code.
In conventional science, schematic diagrams are often used to demonstrated pathways and processes in living systems. What this form of computation biology does is translate an entire such schematic into a series of programming statements that collectively form a computer simulation. The hope is that as this science develops, programs could automatically translate such schematics into simulations.
The advantage of such a method is that it can assess, reaffirm and predict sequences and events in biological systems, with an emphasis on logic over hard numbers.
Jasmin Fisher and her husband, Nir Piterman, have literally married computer science with molecular biology using this approach, as described in last week’s issue of Nature.
Using software that was originally developed to find errors in microchip circuitry, Fisher and her colleagues — in partnership with Piterman — have traced flaws in signaling pathways in the worm, C. elegans.
A series of computer code corresponding to the pathways was generated, and “standard model-checking algorithms” routinely used in computer hardware were used to determine if the logic in the series of statements is consistent. This makes the information obtained qualitative as opposed to quantitative, since the reliance is on rules rather than numbers.
Any discrepancies in the system would be revealed by errors determined by the code. For instance, among permutations of 48 known mutations that affect vulval development in the worm, only 4 were found to be abnormal based on the program. Two of the four were seen to produce previously uncharacterized effects upon further simulation. These findings were then followed up with laboratory experiments.
While the results of computer programming may not be good standalones for the detection of biological effects, they certainly seem like a great way to manage large volumes of data that are produced in genetic studies, many of which are unmanageable without computational biology. Such programming can allow the distilling of large volumes of data and predict future experiments to reinforce hypotheses.
Comments are closed.