Portal de Periódicos da CAPES

Dark matter

2010; The Company of Biologists; Volume: 123; Issue: 21 Linguagem: Inglês

10.1242/jcs.080317

1477-9137

Mole,

Genetics, Bioinformatics, and Biomedical Research

Howdy, pardners. Here I am in the desert, amid the sagebrush and the cactus and the cowboy saloons (yee haw). Listening to biochemistry lectures (wild Western blots) and thinking about how complicated it all is. So many signals and molecules, like stars in the desert sky. Hard to wrap my head around them. Stars, I mean. And molecules.Like so many biomedical science types, I have physics envy. These amazing physicists, whose brains can comprehend the very, very small and the impossibly big, and the arcane mathematics that describe these things. Green with envy. Okay, only sometimes, like when I'm sipping prickly-pear margaritas under the desert sky.In the post-genomic, post-proteomic, post-interactomic era (now-ish), we have the notion that, despite all the complexity, we should be able to do what the physicists do: input all the variables we can (and even sometimes do) measure, and create models that are predictive of what will happen if we do this or that. Maybe even better and, to a great extent, more than I (for one) might have predicted, this seems to work. We have a name for this sort of analysis: computational biology. Heady and exciting stuff. Because by doing this (I mean, those who actually can do it; not me – I have trouble setting up e-mail), we learn whether what we think we understand actually explains the things we can observe.This isn't a new idea. Well before we had computers, the formalization of biological concepts already had a rich history. J. B. S. Haldane and R. A. Fisher fully grasped how the Mendelian principles of genetics rediscovered in the early 20th century could be integrated with Darwinian natural selection, and invented population biology. W. D. Hamilton and John Maynard Smith, to name only two (but a very brilliant two), developed these ideas to create theoretical landscapes of population dynamics. And as we dive from interacting groups to the individual organism, no less than Alan Turing (who ushered in the computer age with John von Neumann) formalized the ideas of gradients of diffusible signals as controllers of developmental patterning, with some fantastic ideas from Waddington (canalization). Ever deeper into the fundamental workings of life, models of molecular control of gene expression and its chemistry abound and abounded, as these mechanisms became understood. Even the astonishing physicist Schrödinger (of the famous cat) weighed in on these issues.But it is with the completion of the genome-proteome-interactome-etceterome that we have the real potential to formalize the very stuff of life and reduce it to equations, laws and models (if not in that order). At the very least it can tell us whether we're missing something essential.Right?Well, I thought so, in my excitement for this exciting area of exciting science. That is, until I met up with my friend Weasel. You remember Weasel, I hope – he's the one who is scary smart and very funny, and who I love when I'm not hating him for being so smart. (But I'm funnier. I might also be better looking, but really neither of us is very good looking – hey, we're academics.)Weasel and I meet whenever we can, and on this particular occasion our venue was no less than The Eagle, the pub where, famously, all of modern biology began (or at least, the bits having to do with the positioning of nucleotides in DNA and related phenomena, which ain't bad). It was a cold winter night and there was a roaring fire to warm our old academic bones, far away in time and space from the desert where I am sitting now. (With no fire present, or wanted, in this climate, thank you.)Weasel and I spoke of many things, as we do, and the subject turned to this idea of taking biology to the level of testing our models in rigorous, formal ways, to see whether we actually understand what we think we understand. But Weasel wasn't having it at all."Look at astrophysics," said Weasel, after whetting his whistle with a frothy beverage. "They carefully calculated the mass of the observable universe, applied the models of gravitation and concluded that there is something missing – dark matter – and they even know how much dark matter there has to be."I looked at my empty glass. "Dark matter. Is that something to do with the Eclipse trilogy?" I procured another round."No," chuckled Weasel, "that's vampires. Dark matter is mass that isn't detected by particle emission or reflection.""But vampires don't have reflections. How does that work?""Forget vampires." Weasel was on a roll. "The point is that the modelling of the universe doesn't work with what we can observe by conventional approaches. The discovery of dark matter has launched a research initiative by revealing a flaw in the data, based on available – and rigorous – models.""Okay, that's cool," I countered. I couldn't see where he was going. "So?""So," said Weasel, moving his chair a bit nearer the fire. "Ever since we've known about the relationships between DNA and genes, we've furiously modelled complex networks of transcriptional regulation, protein turnover and cellular functions. The models all worked pretty well. We seemed to actually be understanding biology!""But that's the point!" I cheered, drawing looks from our neighbours. "We are making progress and the models reveal how the complexity of the systems explain the biology. What's wrong with that?""Dark matter. Nobody found any dark matter. It all seemed to be on track: genes make proteins, proteins regulate genes. Nothing huge and important was missing. The modellers didn't warn us and say, 'Hey, there's something big that we can't account for'.""Maybe there isn't?" I needed to think – what was I missing? I took a long sip – more of a gulp really."Come on," urged Weasel, "what about microRNA? Hundreds, maybe thousands, of microRNAs that influence protein production and therefore the relationship between gene and phenotype. A huge regulatory component with fundamental roles in biological phenomena.Nobody even guessed that there had to be something like that. All the models and calculations, and nobody guessed. More to the point, nobody told us there had to be something like that, something big we weren't taking into account."Yeah, I got it. "MicroRNAs should have solved a dark-matter problem in biology. But the models didn't point to such a problem. They were found by reduction, not deduction.""So, modelling of gene control and other cellular mechanisms is an exercise. It doesn't tell us what we're missing. I'm not sure what it does tell us. But I am sure I need another pint." The discussion wandered off from there, into vampires I think, but as with so many of my discussions with Weasel, the gist stayed with me.To be fair, my friends who do computational biology are probably more sceptical than I am (or even Weasel, really) about the usefulness and limitations of their approaches. My friend, Professor Frog, who is a brilliant modeller of complex biological systems, often tempers my excessive enthusiasm for his approach (to be exact, he stomps on them), warning of the caveats that he understands far better than I do. He agrees with Weasel's dark-matter argument, with a bit of a catch.We can't stop dreaming of the day when our studies on molecular interactions in biological phenomena are routinely integrated into rigorous computational models. For this to happen, of course, we have to move beyond our simple 'X binds to Y' analysis to determine kinetic parameters of quantified molecules interacting with each other in individual cells, and their consequences over time. Okay, I never said it was going to be easy. But we've started to do this and when we do this more we can move from simply believing stories, because they make 'sense', to a point at which the rigorous analysis of complex systems can hope to tell us that something is missing. Professor Frog doesn't deny that we didn't realize that there is dark matter in our biological universe, but we never looked for it with the rigor needed. He and his colleagues are getting there. Maybe some day, they can give us some idea of how much dark matter there still is to find.But for now, I'll satisfy myself with scanning the desert skies where the astrophysicist's dark matter lurks and might contemplate one more margarita. Soon, I'll be mosey-ing to my campsite (okay, it's a quite nice hotel room), but not just yet. Ooh, was that a shooting star?