Analysis of 23andme by a Genotyping Expert

September 14, 2007

The second biggest blunder in Google’s history, only to the Google Jet having access to a private Federal runway, is the 23andme conflict of interest.

Conflicts aside, it’s always helpful to know how what these genome startups are doing, so I asked an expert in the field of genotyping and genetics (who is open to new consulting projects).

How hard is it to do genotyping via SNP?
It isn’t hard at all until you try to do a lot of it.  In research, you look at SNPs pretty regularly and it is very easy when looking at one at a time.

Trying to look at all the SNPs in a genome at once is a lot harder. You couldn’t possibly look at them all so you use informatics tools to define informative sets of SNPs then ignore the rest. This still leaves many many SNPs to look at so this is cutting edge technology.

What’s the benefit to knowing this genetic information?
Countless things you can do with this info.

First of all this is the beginning of personalized medicine. Once we have this information for large numbers of people we will be able to figure out why one medicine works for one patient and not for another and doctors will be able to better treat people with this information.

That is the good stuff but you will also be far better at predicting possible health problems, which is great for preventative medicine, but is scary if your health care provider gets this info and decides to charge you more or not cover those things you are at risk for.

Another interesting but dangerous application is precisely defining peoples ethnic makeup. Once we know this information for many populations of people we will be able to say for example that someone is 18% Scottish 42% Irish 27% German and 13% English. Cool, but what if someone decides to descriminate against you because you aren’t 100% caucasian? And there are many more things to do with the data as well, no wonder Google and Genetech are involved.

Has the human genome been mapped?
Yeah that has been done for about 5 years with tweaks every now and then. It was originally defined for about 6 individuals. The original effort took A LOT of work and money. Technology has been improving and price is coming down quickly. The cost of DNA sequencing/genotyping roughly follows Moore’s Law.

What’s a genome?
Yeah every organism has a genome. Every cell has the same DNA as all the other cells in your body. So since we know the entire genome for the 6 people I mentioned above we can just look at the SNPs for everyone else and merge the two to know each individuals genome.

How much does this cost?
In a lab it could be done for $2 per SNP. I would guess 23andme is looking at 10s of thousands of SNPs, so the lab price is $50,000. There is a huge economy of scale here with new equipment and custom software for processing and analytics, but their price will likely be very low.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: