Still working on an Interesting piece of journalism by the Grauniad
They ordered a 78 "letter" piece of DNA for a smallpox envelope protein.
As they note, the actual genome is rather longer than that, but as they also note, there are techniques for reconstructing genomes from DNA pieces.
I don't think anyone will be synthesizing smallpox from less than 100 base length pieces, got to be easier ways to do it (start from a chickenpox virus and edit it?); but they make an interesting if overhyped point.
Update: Nick at Scientific Activist also comments.
He notes that this suggests a need for more government regulation.
On reflection I think this is right, not because companies that sell custom DNA sequences on spec are evil or reckless, but because regulation, with moderately draconian penalties, will force honest (and most competent) companies to exercise due diligence (at some cost, but screening custom sequences against listed pathogens can't be that high an effort). As the grauniad notes, the company they dealt with never even thought to check.
Actually synthethising something like smallpox de novo from short custom sequences seems unlikely, if terrorists could do that, why would they bother rather than doing something really nasty.
There might be an issue in this expediting the editing of harmless virii like cowpox into something nastier, change some proteins by adding sequences from worse virii, But, I'll leave it to my betters to speculate on that.
Interesting, like living in the preamble to an Egan novel.
DNA is getting to be too easy.
Of course a Bondian level Evil Company would be indefensible - the DNA is just letters, there's no safeguarding against what they actually send as opposed to what they say they sent in accounts.
But, a rational Evil Company would not expedite the synthesis of a war virus for any amount of money, because anyway it goes, they're dead and have nothing to spend their ill gotten gains on. Right?
Comments
That's interesting stuff, and I wrote something about it too here, although you beat me to it!
Posted by: Nick Anthis | June 14, 2006 8:57 AM
I suppose iit would be easy enough for companies who make oligonucleotides to run a BLAST search on their sequences. But, I think someone determined enough to synthesize a 185 kb genome from oligos could find a way to get around it.
Posted by: EMC | June 14, 2006 3:32 PM
Hello Steinn,
Your proposal for government regulation might have the unintended consequence that a company's lawyers might suggest that the safest policy is not to send DNA sequences to individuals or other minor customers. They might suggest that the profits are low, and the danger of being zapped by the government incalculable.
John
Posted by: John McCarthy | July 5, 2006 8:22 PM
Hey John
I agree, that is a possible outcome, although in practise I'd expect a large fraction of the market for DNA synthesis like that to be a "long tail" - if you're big, you make it in house anyway.
I don't know a solution, relying on company honesty is not the problem, more to make sure they exercise due diligence in not inadvertently shipping pieces of biowar agents.
Since government is likely to take interest, this seemed the most innocuous and effective method to make sure they check.
The danger is not so much the government zap, as actually inadvertently providing resources for a mass casualty attack on themselves.
Tricky.
Posted by: Steinn Sigurdsson | July 5, 2006 10:18 PM