dna as soon through the eyes of a coder

Posted 1 May 2002 at 20:29 UTC by ahu Share This

'If you are a hammer, everything looks like a nail'. Ever wondered how DNA can be a 'blueprint'? It turns out that DNA is more like code than you may have imagined.

For the past years I've been learning more and more about DNA and how it works. Some time ago I discovered that I was wrong in separating my computer experience from my DNA studies.

The more I read the more I saw similarities - which seemed more or less pointless until I realised that this could be a way to instrument codert in rapidly gaining interest and understanding of the workings of DNA.

So if you want to know more, or even better if you DO know more and want to tell me I'm wrong (which is quite likely), please visit my Amazing DNA site.

US invented DNA for DNS, posted 1 May 2002 at 22:14 UTC by badvogato » (Master)

US invented DNA for DNS and it thinks that US should remain its stronghold on root servers: 10 out of 13. It hammered Kashpureff down, when AlterNIC plays with DNA of InterNIC's DNS. Vixie is extremely mad when Dr. Jon Postel remained his posture as that of a great man before the 'i may grow old but i refuse to grow up' playful lad.

Most recently, one UK dumb arse bet that "By the end of 2012, more than 50% of the root servers on the internet will be located outside the United States". To that, our lady says it's impossible since the number of root server is limited by the size of UDP packets. To increase 13 root servers to a number that can be distributed evenly across geographical space requires new DNA for Domain Name Service. a Win-Win situation is hard to gain without political struggles and possibly a world war III.

DNA for DNS? badvogato gone bad?, posted 2 May 2002 at 00:55 UTC by atai » (Journeyer)

Badvogato, are you bad? Eh, I mean, what are you talking about? What does the DNA article have to do with root servers for DNS?

0xc3, posted 2 May 2002 at 04:35 UTC by tk » (Observer)

I wonder whether there have been any attempts at sequencing the DNA of bacteria, which would seem to be the genetic equivalent of "Hello world"?

There is no shortage of (sequenced) bacterial genomes., posted 2 May 2002 at 11:22 UTC by thomasd » (Journeyer)

You can download your favourites from here.

Some of them are quite small -- for instance, Mycoplasma genitalium weighs in at only 580kb of DNA. That's still a fair amount of information, though. Perhaps `hello, world' statically linked to an operating system kernel...

For a selection of larger genomes (including human), try Ensembl.

who runs PowerDNS?, posted 2 May 2002 at 14:25 UTC by badvogato » (Master)


DNA as source code, posted 2 May 2002 at 15:54 UTC by adulau » (Journeyer)

I really love the analogy of DNA and source code. It would be really interesting to have a finite state machine to take DNA and "run it". But when you see that it seems complex and not so easy... (concept are the same but way of operation is not)

If you can prove the DNA is source code and if you can prove that source code is speech. (Ok you have to prove the too, and second is more complex ;-))

You can prove that DNA is the expression of Nature. Nature can't be protected because there is anteriority on the code itself.

No software patents.... (and dna patents) Just Freedom...

Ok, that's for the long-term.

As time goes by..., posted 2 May 2002 at 17:31 UTC by sye » (Journeyer)

Let Nature do its work!

If DNA is the expression of Nature and only natural evolution works its way in over long period of time and render its grace in all creatures, shouldn't we do the same thing with our source code? Give enough time for source code to interact with all creatures before acting like a God to change DNA at a whim.

No, a second thought, souce code is not DNA but the design principle of your source is. Source code without clear thinking on the divine design is just like your facial hair.

You can't "run" DNA, posted 2 May 2002 at 20:26 UTC by cactus » (Master)

adulau: the problem is, you can convert the DNA into some binary representation, but you can't really build a 'virtual machine' (or, more precisely, 'virtual cell') to run it in, since you can't 'create' the proteins. That would need a simulation of the interaction of proteins of arbitrary amino-acid sequence, both with each other and with the DNA strand.

I don't think you could even model the interaction of two simple proteins.

powerdns, posted 2 May 2002 at 21:21 UTC by ahu » (Master)

yeah, ahu runs powerdns - badvogato did his research! But I fail to see how this is relevant to the DNA discussion.

'running the code', posted 2 May 2002 at 21:24 UTC by ahu » (Master)

This really is rather like 'The Matrix' - once you synthesize the proteins, *nature* executes them. That is the weird bit.
This is all not very practical - you currently need a cell to run the 'code' that runs the cell. Nothing practical about that.
Another thing to worry about.. it appears that there is only in the order of 20 megabytes of DNA involved in building a human.. surely that is not enough!
That's what I'm trying to understand right now.

it is so obvious, posted 2 May 2002 at 21:49 UTC by badvogato » (Master)

It is so obvious. Godzilla could have been smarter than ahu at this point. And allow me to post a comment on your library. You have quite a few books on engineering part of writing code which can be run on any dumb machine but not a single book to reveal the trade secret on how to design the next generation machine that can process sub-proteins. I am watching HOOTS in the night. That was fun!

DNA -> DNS -> DNS! >> ASS is ME, posted 2 May 2002 at 22:05 UTC by badvogato » (Master)

I see! But still fucking with DNA runs the risk of either childless or too many offsprings than one can actually handle. It's much evil than writing program. Watch your back, man. God bless you.

Squeak again, posted 3 May 2002 at 07:13 UTC by tk » (Observer)

badvogato is really bad.

The Squeak project (again!) project may hold the answer to ahu's questions. It is a Smalltalk system which contains, among other things, a Smalltalk VM written in itself. The uncompressed source code of the entire system -- including VM -- stands at 15M.

As for cactus's concern, perhaps organic chemists can provide us with a clue?

"The Cerebral Code", posted 3 May 2002 at 21:32 UTC by badvogato » (Master)

I am convinced that William Calvin has to be related to Dr. Oliver Sacks. I read Dr. Sacks' book: "The Man Who mistook his wife for a hat". One critic says the book is "A provocative introduction to the marvels of the human mind..." <pause 20 second...waiting for a few clicks...>

See, i was right! William H. Calvin is a theoretical neurophysiologist. He is almost like a brother to Dr. Oliver!

Overlapping genes, posted 4 May 2002 at 05:00 UTC by ncm » (Master)

You mention what a cool hack it is that genes can overlap -- offset by one or two base pairs, maybe going in opposite directions. In fact this isn't just (or even mainly) to save space. It's an error detection mechanism. If anything changes in an overlapped sequence it breaks two genes, and that error probably doesn't get propagated because the phenotype is too buggy.

Overlapping helps make a pair of genes "holy".

HIV uses this, coupled with a really buggy copying enzyme, to enable reliable reproduction despite a really fast mutation rate.

Not New, posted 8 May 2002 at 06:07 UTC by BenFrantzDale » (Journeyer)

This idea is covered in some detail in Gödel, Escher, Bach: an Eternal Golden Braid by Douglas Hofstadter. It's an amazing book in general; I highly recomend it.


New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

Share this page