I was recently interviewed on Radio Sandy Springs by Sharon Sanders of FluTrackers. We talked about pandemic influenza H1N1 virus. Listen to the show below.
Download Infectious Disease Hour January 4 2010 (20 MB .mp3, 57 minutes)
Radio Sandy Springs 1620 AM is a low-powered Atlanta-based talk radio station that simulcasts on the Internet.  They broadcast a weekly ‘Infectious Disease Update’ with interviews with clinicians, scientists, researchers, and even historians. You can find an archive of recent Infectious Disease Hour shows here.
I just thought of a way to do ethical human trials
I was just listening to this pocast (I use a program called miro to retrieve all your words of wisdom)
We are real close to completely sequencing individuals now.
Those sequences go in one database.
For the sake of clarity, we keep all the disease genomes in a different database.
Then we link the two by getting as much disease history as we can (that we're certain of, of
course, not just “the flu”). We test all the seropositive reactions etc. As close as we
can, who had which strain.
Not very useful at first, but we just keep adding data as we get it.
Tying one strain to even two different genomes automatically eliminates quite a bit of it,
and accuracy keeps going up from there.
Through the miracle of statistics, we only need about 1000 human sequences to get a pretty
estimator for all of the USA.
Nobody gets hurt except for some needle sticks.
In theory, if you had just one human cell and one virus particle of the type that killed it, you
could sequence them both and create a “deadly” link between them.
At the very least, it indicates getting a very thorough history from everyone getting their DNA read.
It also indicates that if a new disease emerges, “patient 0” should go to the front of the line.
I'm sure someone has thought of all this already, but it seems like an idea worth repeating.
I just thought of a way to do ethical human trials
I was just listening to this pocast (I use a program called miro to retrieve all your words of wisdom)
We are real close to completely sequencing individuals now.
Those sequences go in one database.
For the sake of clarity, we keep all the disease genomes in a different database.
Then we link the two by getting as much disease history as we can (that we're certain of, of
course, not just “the flu”). We test all the seropositive reactions etc. As close as we
can, who had which strain.
Not very useful at first, but we just keep adding data as we get it.
Tying one strain to even two different genomes automatically eliminates quite a bit of it,
and accuracy keeps going up from there.
Through the miracle of statistics, we only need about 1000 human sequences to get a pretty
estimator for all of the USA.
Nobody gets hurt except for some needle sticks.
In theory, if you had just one human cell and one virus particle of the type that killed it, you
could sequence them both and create a “deadly” link between them.
At the very least, it indicates getting a very thorough history from everyone getting their DNA read.
It also indicates that if a new disease emerges, “patient 0” should go to the front of the line.
I'm sure someone has thought of all this already, but it seems like an idea worth repeating.