Bioinformatics’ anthropic principle
The anthropic principle is a strange creature. In its weak form, it states that we exist, therefore the universe must possess properties amenable to life. This is trivially apparent. The anthropic principle’s strong form, on the other hand, states that we exist, therefore the universe must possess properties amenable to life (Okay, I’m with you so far), therefore the universe was created so it would create us (Wait, what?). I’ve always been puzzled by the strong anthropic principle—the weak form is elegant, for it implies that we simply wouldn’t be around to observe a universe that would not permit our existence. The strong form, however, is creationist poppycock.
Playing with bioinformatics tools has made me reflect: What would the bioinformatics version of the anthropic principle say? It seems a joyous happenstance that bioinformatics tools are largely open-source, that data formats are usually text-based and standardized (if not always well-followed), and that operating systems are almost universally *nix-flavoured. Though I can conceive of the converse, in which bioinformatics tools were exclusively proprietary, running only on Windows, employing obfuscated binary file formats that were undocumented and specific to each vendor, it scarcely seems a world in which I would want to live. Performing analyses by tying together different tools would be impossible—we would need a single monolithic program, undoubtedly driven by an excruiciatingly complex GUI, to perform every task required in a pipeline. Tool advancement would be much slower, given that we would no longer have the incremental advancement made possible by openly published algorithms. Scaling to ever-larger computing clusters would be needlessly complicated by the need to license software. The world, in short, would suck.
Bioinformatics’ embrace of open technologies can be summarized by what I shall call the weak form of the bioinformatics anthropic principle: bioinformatics exists, therefore there must also exist a confluence of computing technologies amenable to it. I believe electronic computation is indispensable to bioinformatics—though one can do bioinformatics without it, this would be tantamount to doing astronomy without telescopes. Bioinformatics requires we perform innumerable repetitive computations to make sense of our data, for which, sadly, our puny human meat-brains are poorly suited. But the wonderful thing about bioinformatics is not simply that it uses computation, but that it has settled on a variety of open means of doing so. Though bioinformaticians make imperfect software engineers (a fact which seeped deep into my marrow when I perused the ChIPPeakAnno source code), they nonetheless adhere to the best parts of the Unix philosophy, which makes the cockles of my heart swell with joy.
But what about the strong version of the anthropic principle? The existence of the weak demands we formualte the strong as well. I propose the following as the strong version of the bioinformatics anthropic principle: bioinformatics exists, therefore there must also exist a confluence of computing technologies amenable to it, therefore the Omnipotent Bioinformatics Creator imposed technology choices in the field’s nascent days so that I would be happy. I believe this bioinformatics strong anthropic principle to be every bit as deductively incontrovertible as its non-bioinformatics counterpart—it’s the anthropic principle, not the anthropic conjecture, right?
This leaves open the question, however, as to who is the Omnipotent Bioinformatics Creator. In this, there can be no doubt—just as a rather famous J.C. often accompanies the non-bioinformatics strong anthropic principle, we have our own supreme mythical being in J.C.V.