Evolution is such a hacker

Transterrestrial Musings links to an article on "The Big DNA Letdown." The sequencing of the human genome was insanely overhyped back at the funding stage, and has produced almost none of the promised medical wonders.

This was, of course, easily predicted. The linked article attributes the letdown to something that wasn't known at the time, "epigenetics," the influence of cellular context on the expression of genes. But it was predictable even based on what was known.

The linked article describes the pathetic state of our understanding at the time: "But in the past few years, researchers have come to recognize how grossly oversimplified was the so-called central dogma of molecular biology — the notion that DNA makes RNA, RNA makes proteins, and proteins make us." The sequence of DNA codons produces a predictable sequence of amino acids in the protein.

And then the protein folds. Sequences of amino acids from far distant parts of the protein are brought next to each other, the chains twisted and bent, to make the active sites of the protein. The protein folds itself in milliseconds. Human beings with supercomputers pretty much can't do it at all. So, knowing the DNA sequence still leaves you  an NP-complete problem with large n  away from anything biological.

I happened to know about the folding problem because I shared an apartment with a couple of theoretical biochemists who were working on it in the winter of 1986-7. It's still far from being solved. Solving it would be one of the biggest steps that could be taken, to make genomic sequencing useful. But as I commented over at Transterrestrial Musings:

Protein folding — being able to predict the 3D shape of a protein from the amino acid sequence — would be a lot more immediate use. But the public mind isn’t prepared to accept “protein folding” as a magical incantation. “DNA,” on the other hand, is used as magic in innumerable movies, TV shows, and comic books. Anything to do with DNA is already overhyped.

The Universe is a computer simulation

Curmudgeonly & Skeptical has an astronomical photo of a "hole in space" and asks what is it? The answer is almost too obvious to require explanation, but I did anyway.

I'm sure all your readers are familiar with the idea that the universe is actually a computer simulation. Every simulation has its glitches. The hole in space is a minor one.

Some of the other obvious glitches: half the events in Spanish history happened in 1492; the Japanese and Scandinavians morphed from Samurais and Vikings into inoffensive Saab jockeys and Hello Kitty consumers; a movie about a mathematician won an Oscar; a Kenyan became President of the US; the Red Sox won a World Series; John Travolta and Mickey Rourke won't stay dead; vuvuzelas; you don't even know who Renee Olstead is; the state of Massachusetts; and the Turing word for this post is "lasouna", which is ludicrous.

Not only is our "universe" a simulation, but it's running under limited resources. Take the way certain mathematical constants get reused in completely different contexts. There's an old joke about an actuary giving a presentation to the executives of an insurance company. He puts up a slide with the formula for the number of life insurance customers who will die in the next year. One of the executives interrupts. "What's that in the formula? That funny symbol?"

The actuary looks and says, "That's pi. You know, the ratio between a circle's circumference and its diameter."

The executive explodes. "What malarkey is this! What does a circle have to do with how many people are going to die!?"

But given that it takes an infinite number of bytes to store pi, you can see where using the same constant for both, really saves on the memory.