Skip to content

Comment on Guy commits his genome to Github, smartass forks and issues a pull request

Comments

shame, I did this a few weeks ago: http://github.com/orta/dna

In your README you state: I find it fascinating that you can get a text file, roughly 25meg big that contains what I'd imagine to be terabtyes worth of data

I just thought I'd point out that you're really just getting a 25MB patch to a 3GB file (which is really duplicated to 6GB).

Ah, now that makes sense, thanks mbreese

I would say you dodged a bullet. This guy's repo is fast becoming a gag: https://github.com/msporny/dna/network/members

Also, he explicitly released it under the Creative Commons Public Domain License. Incidentally, I wonder if that has any patent implications? Does having genomes in the public domain prevent pharma companies from patenting their "discoveries"?

Heh. The GitHubt default text on that page is accidentally funny in itself ;)

msporny created dna and everyone else forked it. This is the family tree.

Aren't you not supposed to be able to patent facts? Seems like your DNA structure should be classified as a fact.

see the BRCA1/2 (myriad) debacle. horrible mess.

The public domain shouldn't matter. Publication could.

Manu said [1] he was the first to do so, but looks like maybe you beat him.

[1] http://manu.sporny.org/2011/public-domain-genome/

So, where/how do you get a dump of your DNA, anyways? I suppose it must cost a pretty penny, too?

He used 23andme. And they only give you data for around a million SNPs (http://en.wikipedia.org/wiki/Single-nucleotide_polymorphism).

The cost of sequencing decreases by a factor of 10 (!) a year per base pair. I will wait until it's cheap to sequence my whole genome, before I'll shell out the money.

(Synthesizing also gets cheaper by around the same factor. But it started from a higher level of cost.)

I saw the XMas sales for 23andme and I thought the money at that time was well worth the knowledge.

If and when they offer the service for a complete sequence, I'll probably pay the money again, just to have a complete record.

Did you find anything interesting? Ailments or ancestry?

I found all of it interesting, but I'm a bio geek. I'm also planning to have kids with my fiancé so it's also cool because of that.

Some highlights:

Both my fiancé and I are carriers for a mild form of Hemochromatosis, so we have a 1/4 chance of having a kid with it.

I am a little bit Asian! That's probably the Jewish ancestry. My fiancé thought he had a black ancestor but it turns out was 100% European. My best friend, whose father had always told her that her grandmother was Native American, found out he was lying to her (which did not surprise but did disappoint her.) Also the heat map of where your mitochondria are from is fun; it confirmed what I knew already but it was still really cool to see that the first migration of my maternal line out of Greece was with my grandmother.

But that's really just the tip of the iceberg. Part of what makes 23andme so fun is you get to learn a lot about biology from the framework of your own genes. Each SNP they report on has a bunch of information associated with it. It's only boring if you aren't interested in biology.

That sounds cool. Can they tell you from which European ethnic groups you come from too? How specific are they about this? Do they say French, German, etc...? Or do they say South of France Celtic, Germanic tribe from the North and so on?

Right now the only categories where they give you your percentage and paint your genome accordingly are European, Asian, and African.

The cool thing about 23andme though is that they're always rolling out new stuff. There is a separate, more detailed ancestry break-down in their "labs" (a part of the website for experimental stuff) but it's not very good. Part of the problem is that there isn't enough research pinning down large swaths of the genome to more specific areas. So what 23andme does is use user reported data- which is problematic due to globalization and limited by people's knowledge.

The only other things that are good and can really pin down geography are your mitochondrial DNA, like I mentioned, and the Y chromosome.

Aaaah, not very useful for me then. I live in Europe and doubt that I have any Asian or African ancestry. Maybe some Middle-Eastern or Sephardic Jew but I guess that falls in the same race as Europeans so they won't be able to tell me. $260 was going to be a big investment for me, I will wait until they can tell the difference between early Europe's ethnic groups: Celts, Germanic tribes, Iberians, Basques, "Roman", etc... They need to study people living in Europe in kind of isolated communities, remote mountains or whatever, then they will be able to tell the difference.

And even then it may be hard. There has been mixing up even back then.

I didn't, I found the 23andme website pretty boring, infact, it basically turned into a $200 git clone joke I made a few weeks ago

Any predictions of when sequencing the whole genome might cost around $200?

Just take the current price, and extrapolate. Not more than a few years for researchers. I don't know how retail will develop, since there may be regulations and other overheads.

For my friends in systems biology it's now almost cheaper to just send their microbes in for sequencing than doing their own gel electrophoresis.

(If you want to read around in Wikipedia, also have a look at Southern blots and Western blots, and DNA microarrays.)

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.