Color me “skeptical”


I stumbled across this article today about an application in DNA research of a methodology commonly used in gravitational physics:

“‘Thanks to the Human Genome Project, biology and medicine today may be at a point similar to where physics was after the advent of the telescope,’ said Orly Alter, assistant professor of biomedical engineering at the university. ‘The rapidly growing number of large-scale DNA microarray data sets hold the key to the discovery of cellular mechanisms, just as the astronomical tables compiled by Galileo and Tycho after the invention of the telescope enabled accurate predictions of planetary motions and, later, the discovery of universal gravitation. And just as Kepler and Newton made these predictions and discoveries by using mathematical frameworks to describe trends in astronomical data, so future discovery and control in biology and medicine will come from the mathematical modeling of large-scale molecular biological data.’

In a 2004 paper published in the Proceedings of the National Academy of Sciences in collaboration with the late professor Gene H. Golub of Stanford University, Alter, who holds a Ph.D. in applied physics, used mathematical techniques inspired by those used in quantum mechanics to predict a previously unknown mechanism of regulation that correlates the beginning of DNA replication with RNA transcription, the process by which the information in DNA is transferred to RNA. This is the first mechanism to be predicted from mathematical modeling of microarray data.

[…]A DNA microarray is a glass slide that holds an array of thousands of specific DNA sequences acting as probes for different genes, making it possible to record the activity of thousands of genes at once. Making sense of the massive amount of data DNA microarrays generate is a major challenge. In her Genomic Signal Processing Lab, Alter creates mathematical models by arranging the data in multi-dimensional tables known as tensors. She then develops algorithms to uncover patterns in these data structures, and is able to relate these patterns to mechanisms that govern the activity of DNA and RNA in the cell.”

Read the full article here.

Here’s the thing: even though Einstein’s tensor equations have been well known for decades, there are still only a few known solutions. And those solutions are only valid in special, almost trivial cases. The Schwartzchild solution describes a motionless point singularity. The Kerr solutions describe the same singularity but adds in a degree of angular momentum.

Pretty much anything more than that and the equations get too complicated for us to be able to solve in closed form. We can find a solution to some circumstances by a brute force numerical calculation, or by ignoring higher order effects, but then all we end up with is an approximation and not really and understanding.

The idea that we’re going to be able to take the same techniques and apply them to the much more complicated situation of the genome strikes me as just a bit over optimistic. (I’m guessing it’s the reporter’s assertion and not Alter’s.)

Besides, while we can do a pretty good job of hitting Saturn with a probe launched from Earth, we still include burners to tweak the system in flight.

And don’t even get me started on the anomalous slow down that seems to be observed in the Pioneer probes. Either the higher order terms in the calculation are more significant than people have thought – or there’s something totally unexpected going on. Which is okay in a probe far away from earth. Discovering the same sort of unexpected behavior in e-coli being modified to produce hydro-carbons and unexpectedly getting something “very not good” presents a more serious problem.

I guess I’m a little reluctant to just start messing around with things that are as apparently non-linear as biological systems appear to be. Smacks of playing God and all that. Which bothers me for a whole host of other reasons…

The Author

Episcopal bishop, dad, astronomer, erstwhile dancer...