Software Lets Scientists Assemble DNA

Velcroman1 writes "Biochemical engineers can now download a piece of software and with a few simple clicks, assemble the DNA for new life forms through their laptops. 'With the proper computer tools, biologists can write their own genetic code — and then turn that code into life,' said biochemist Omri Amirav-Drory, who founded Genome Compiler Corp., the company that sells the software. He demonstrated at a coffee shop early one morning by manipulating a bacteria's genes on his laptop. The synthetic biology app is still in beta; on Jan. 15, the company added an undo feature and support for new DNA file formats. Building creatures is increasingly like word processing, it would seem. But such is the strange reality in the age of cheap genome sequencing, DNA synthesizing and 'bioinformatics.'"

Comments

Think I've played this game already

[dywolf]

It was called Spore.

Sounds great

[emagery]

DNA is a programming language after all... but knowing the character set is far from understanding the foibles of the programming language itself. We need to have a deeper and more complete understanding before distributing this kind of power.

Undo?

[meta-monkey]

"Undo feature?" Shouldn't the command to eliminate your unwanted DNA creations be called "Abort?"

Misleading title

[subanark]

Wow, that is really a misleading title for those in the field. "Assemble" generally refers to solving the jigsaw puzzle of putting digitized DNA fragments generated from a sequencing machine together to form contigs which can eventually be assigned to a chromosome.

See:
http://en.wikipedia.org/wiki/Sequence_assembly

Re:Misleading title

[Hythlodaeus]

The whole post reads like a bio version of "And then I cloud-sourced an internet to my giga-drive!"

Re:So -- the terrorists win in the end

[ArcadeMan]

So what you mean is that we should be monitoring Linux users?

Re:So -- the terrorists win in the end

[micromoog]

QUICK, STOP ALL SCIENCE

Re:So -- the terrorists win in the end

[Baloroth]

It takes a special kind of terrorist to deploy a bioweapon, because bioweapons don't distinguish based on religion (although you could theoretically make one that distinguishes on race, it's a bit tricky). That means it'll hit everyone indiscriminately, and not even most terrorists want that. The only ones who would use something like that are people who want to destroy everyone, and finding a large enough group of people (as you would need to create and deploy such a weapon) willing to do that is quite difficult.

Also even the most lethal bioweapons won't kill everyone, whether thanks to natural immunity or proper quarantine procedures, a lot of people will survive. Anything nasty enough to actually kill everyone will almost certainly burn out very rapidly.

Sussman: Emacs mode to edit genome

[ODBOL]

In the late 1980s or 1990ish, I attended a meeting sponsored by the National Science Foundation, to promote interaction between biologists and computer scientists. Much of the discussion focussed on designing algorithms and producing programs to answer questions posed by biologists. That part of the discussion was dominated by laments: biologists describe problems, computer scientists create programs to solve them, biologists find that the solution isn't really what they wanted.

Gerald Sussman (MIT, creator of Scheme) was at the meeting. At one point he got excited, and captured the podium. Alas, there is no transcipt, but here's my paraphrase of his inspiring speech:

Writing programs to serve biologists is cool as far as it goes, but our collaboration should cut much deeper. The genetic code is a programming language, and we should help biologists figure out the structure of the programs written in the alphabet of the bases. What I really want is the Emacs mode to edit the genome, so I can give myself a prehensile tail.

I have a great memory. I remember good stuff, and some of it happened. Please don't blame Mr. Sussman for any idiocies in my paraphrase. Maybe I projected the prehensile tail from my own repressed desires. But, I do think Mr. Sussman deserves great credit for observing the deep conceptual connections between CS and genetics at a time when very few of us thought beyond the idea of writing computer programs to help solve genetic problems.

Re:Protein expression?

[CanHasDIY]

Wonder what the output device will look like?

Genitalia.

epigenetics?

[DdJ]

Does the tool let people specify various epigenetic factors, such as methylation? This is a thing that's pretty important, but that a lot of people don't understand well (and some refuse to believe there's anything to understand there).

If so, wow.

If not, this is going to have some severe limits in utility. Useful, certainly, but completely incapable of producing working DNA for, say, a human being or a giraffe.

Re:So -- the terrorists win in the end

[ultranova]

And how long will it be until extremists design and assemble a lethal and unstoppable virus this way and trigger a global epidemic that wipes out humanity in the name of Allah?

Probably forever, because:

1. Wiping out humanity is the one thing anyone - including the extremists - ought to understand is guaranteed to royally piss off any creator god that might be behind human existence (or any being even remotely interested in humanity, for that matter).
2. Politically motivated terrorism doesn't exactly have many scenarios where actually ending the world would get you what you want either.
3. It's pretty hard to imagine that fundamentalists could outsmart biologists who, after all, also have access to this tool to make a cure.

Nice work, Omri; you've just handed them the tools.

On the other hand, idiots who think other people are cartoon supervillains and appeal to that caricature to argue against new tools are certainly capable of killing millions by hindering the War on Disease. You and everyone who modded you up ought to be ashamed of yourselves. You're just as bad as the anti-vaccine people, except you don't even have misfiring parental instincts as an excuse.

Scare Quotes Not Needed

[me01chanl]

It's no more appropriate to say "bioinformatics" than it is to say "algebra" - they're well defined fields.

not so ubiquitous

[pepty]

The article breezily mentions creating a genome from scratch, but it's not really that easy. Say you wanted to use the services mentioned in the article to create smallpox, a 186 kb (kilobase) virus, from scratch. Genome compiler software would be a way to design the project on a computer, but that's about it. The services mentioned in the article will certainly synthesize oligonucleotides into genes (100 to 10000 base pairs) and put those genes or operons into vectors and ship them to you, but building a whole virus would be a long involved project and would get special attention. Even having them just make parts of smallpox genes would probably throw up red flags in the software; it would be pretty trivial for checks like that to be automatic. But say you get your genes or operons in the mail. You would then need to assemble all of those bits into one genome. That involves a lot of intermediary steps of cutting and pasting, replicating (first by a PCR machine, then in host organisms when the pieces get too big for PCR), purifying, and then cutting and pasting again. Fairly standard molecular biology, but harder with such long pieces of DNA. Then it's off to the biosafety level 5 lab to package the DNA into chickenpox viral capsids or find some other way to get your viral genome into human cells intact. Then you'll need to culture the virus in (presumably) human cells, yet another skill set.

All in all you would need to have access to a lot of the equipment and skills found in molecular biology and virology labs to get the job done, not just mail order DNA.