A biological computer: wouldn’t that be cool. It wouldn’t need wires or batteries but could perform all the logical operations of a digital computer. Bioengineers have succeeded in manipulating genetic machinery to perform most of the operations of Boolean logic — that is, mathematical functions using only on or off states, like 1’s and 0’s. Since computers only recognize 1’s and 0’s, Boolean logic is a good fit for electronic switches.
Named after George Boole, a 19th-century mathematician, Boolean logic is the basis of digital circuit design. For example, the AND function means that both conditions must be on to trigger the circuit. In notational form, where 1 means present and 0 means absent,
1 AND 1 = true or “on”
0 AND 0 = false or “off”
1 AND 0 = false or “off”
In the OR function, either one (or both) of the inputs can be on to trigger the circuit:
1 OR 1 = true or “on”
0 OR 0 = false or “off”
1 OR 0 = true or “on”
Negation of these functions, called NAND and NOR, reverse the outputs. For a logic gate with two inputs and one output, such as a transistor, there are 16 possible operations (AND, NAND, OR, NOR, XOR, XNOR, …). All but the last two, XOR (exclusive OR, true if the inputs are different, but false if the inputs are the same), and XNOR (its negation), had been engineered in bacteria, but those two were more challenging.
Now the journal Science has announced the implementation of XOR and XNOR in bacterial circuits by a team at Stanford and another team at MIT. With the logical operator set complete, the way is paved for more advanced digital signal processing using biological circuits. Networks of biological circuits could be used to “study and reprogram living systems, explore biomolecular computing, and improve cellular therapeutics,” according to the Stanford team.
Both teams were able to manipulate and regulate the inputs and outputs of DNA transcription, using the readily available enzymes in bacteria (promoters, terminators, recombinases, integrases and other cofactors). The presence or absence of the machines determines the output — the amount of messenger RNA produced for a given gene. In bacteria, with the DNA arranged in plasmids (circular loops), the ingredients are easier to manipulate: add the ingredients (input) and measure the mRNA produced (output). The Stanford team called their transistor-like mechanism a “transcriptor.”
The team realized that they were simply altering an already-existing biological logic. “Organisms must process information encoded via developmental and environmental signals to survive and reproduce,” they said (emphasis added). They recognized, further, that their operations are trivial compared to what living organisms do. “Researchers have also engineered synthetic genetic logic to realize simpler, independent control of biological processes.”
In the same issue of Science, Yaakov Benenson, from the Swiss Institute of Technology, agreed that living organisms use recombinatorial logic:
Logic gates evoke images of circuit boards, but cells are arguably equally good in relying on logic computations. A classic example is the Lac operon, which activates itself upon the condition “lactose AND NOT glucose”. In recent years, there have been multiple reports on rationally designed, genetically encoded logic gates and circuits in living cells. Just like the Lac operon, these gates receive two or more molecular signals (inputs) and generate a product (output) whose level is logically linked to the inputs.
What this implies is that bioengineers are not simply using cells to do what they would not otherwise do. They are not, so to speak, pushing a herd of cows to move through artificial gates that humans arranged to act like switches, but rather attaching milking machines to the cows to take advantage of existing functions for their own desires. They are employing “rational design” (a synonym for intelligent design) to steer existing functions for human goals. Arguing from the lesser to the greater, it must have taken rational design to bring the cell’s logical operations into existence.
Benenson believes that “natural evolution” had a harder time encoding the XOR and XNOR operations, but for that assertion, he passed the buck to Leslie Valiant of Harvard in the references. Valiant’s 2009 paper in the Journal of the Association for Computing Machinery seems highly speculative. Utterly dependent on the validity of natural selection, Valiant’s theory (or rather “suggestion” or “notion”) starts with an embarrassing admission about Darwin’s theory:
Darwin’s theory of evolution suggests that such mechanisms evolved through variation guided by natural selection. However, there has existed no theory that would explain quantitatively which mechanisms can so evolve in realistic population sizes within realistic time periods, and which are too complex. In this article, we suggest such a theory. We treat Darwinian evolution as a form of computational learning from examples in which the course of learning is influenced only by the aggregate fitness of the hypotheses on the examples, and not otherwise by specific examples. We formulate a notion of evolvability that distinguishes function classes that are evolvable with polynomially bounded resources from those that are not…. We suggest that the mechanism that underlies biological evolution overall is “evolvable target pursuit”, which consists of a series of evolutionary stages, each one inexorably pursuing an evolvable target in the technical sense suggested above, each such target being rendered evolvable by the serendipitous combination of the environment and the outcomes of previous evolutionary stages.
It should go without saying that Darwinian evolution has no targets to pursue. It is not trying to “learn” anything. Maybe that’s why Benenson makes only a passing reference to this paper, and the Stanford team doesn’t mention evolution at all. They do, however, speak of “natural systems” and “natural operons” — the language of computer design.
Our uniform experience with computer circuits that use logic operations is that they originated from intelligent causes. The inference to the best explanation for biological circuits that employ logic operations is that they, too, are the work of intelligent causes.
It doesn’t matter if the Stanford team and commentator Benenson “believe” that bacterial logic somehow evolved by natural selection, despite the problem of getting “mechanisms…which are too complex” to evolve “in realistic population sizes within realistic time periods.” They can choose to believe that if they want. What matters is the logic behind their work. In this case, they are implying, “Do as I do, not as I say.”