Solving the genome puzzle(plus.maths.org)
plus.maths.org
Solving the genome puzzle
http://plus.maths.org/issue55/features/sequencing/index.html
12 comments
You know, I never really thought about it before, but there really isn't a whole lot of data in the human genome. There are approximately 3B base pairs in it, which sounds like a lot, but think of it in terms of binary data. DNA is composed of 4 bases, so that's base 4; converting to base 2 gives you approximately 6 billion bits of data. While that's a lot of data to sift through by hand, it's not a lot of data for a computer to deal with. Obviously, there's a lot we simply don't understand yet, but I wonder how much your average hacker can bring to the table. After all, That's only 5.5GB to toss around -- nothing, these days.
> That's only 5.5GB
6 billion bits = 0.75 GB. I wonder how compressible it is, on top of that.
Another thought: two humans are 99.9% identical, so a diff is only 0.75 MB. Multiply by an estimated 6.79 billion world population. 5 billion MB = 5 petabytes. Google has enough hard drives to store the firmware to every human being in the world.
6 billion bits = 0.75 GB. I wonder how compressible it is, on top of that.
Another thought: two humans are 99.9% identical, so a diff is only 0.75 MB. Multiply by an estimated 6.79 billion world population. 5 billion MB = 5 petabytes. Google has enough hard drives to store the firmware to every human being in the world.
Well.... proteins play an important role in how DNA is expressed. And you get your initial set of proteins, like your DNA, from your parents (mom mostly), so they set the starting conditions. Then during your life what you do has a methylation effect on your DNA. Thus the total information that makes up a human is actually quite a bit bigger then the information contained in his or her DNA.
> I wonder how compressible it is, on top of that.
You can download a copy from Wikipedia and get back to us :-)
You can download a copy from Wikipedia and get back to us :-)
There's tons of metadata on top of this. Organization and patterns we haven't figured out yet. It's like saying a Linux distro is 700MB-- there's a lot of machinery in there.
Oh, of course. It's just amazing to me that it's so small but does so much. I do wonder what one could do with it, though, without a serious background in bioinformatics.
Complex-looking fractal structures can be described with a simple formula, with very small amount of data and fractals are everywhere in nature. For example http://en.wikipedia.org/wiki/Barnsley_fern
For those who find that interesting, read up on L-Systems (http://en.wikipedia.org/wiki/L-system). There was a great free ebook on this a while ago, but unfortunately the title (and URL) escapes me at the moment.
5.5GB in one cell. The human body in a way is like a massive parallel computer, with cells giving and receiving signals and affecting other cells. There are approximately 30 trillion cells. Although most of the DNA is junk, the genome is better compared to software (containing data + algorithms).
This is awesome, I've encountered this bioinformatics stuff before in a cluttered textbook and didn't follow it. Reading this at least gave me a grasp of the knowledge. Should come in handy when it gets introduced in my course.