Which genes should I get sequenced?

December 16, 2011 3:12 PM

Which of my genes should I have sequenced, mostly for an art project?

I have the resources to get stretches of my DNA sequenced. I want to generate some DNA sequences for use in a slightly odd "self portrait" art project.

So, if you were me, which gene(s) (or non-coding sequences) would you want to get sequenced?

My rough criteria are:

(a) Some variability between individuals; I'd prefer not to end up with a sequence identical to one I could've downloaded from NCBI.

(b) Encoding a known (and important/intresting) gene or non-translated region.

(c) Not expected to have too much in the way of complex structure (hairpins, etc)

My initial ideas are things like: A region of my HLA gene(s) and/or stretches of transcritionally active endogenous retroviruses (HERVs), which I assume aren't too strongly conserved. I'd prefer to stay away from anything too medically relevant, e.g. not Huntingtin.

Bonus points if you know the gene symbol and/or sequences for pre-designed primers, but no worries. I'm a molecular biologist but not a geneticist, so go easy on me!

posted by metaBugs to science & nature (6 answers total)

Yeah, HLA was my first thought too.

I'd probably throw in a few genes known to affect physical appearance, such as MC1R I'm sure there are some fun arty things you could do with that.
posted by grouse at 3:37 PM on December 16, 2011

You don't give too many details about the portrait you are trying to make but if I were doing this, I would want a mix of genes that were sure to be different between individuals (different humans) and then at least one of the genes that is amongst that very small amount which is different between humans and chimpanzees. And then maybe something that humans share homology with in yeast, etc. I couldn't tell you names of such genes offhand, but someone else can?
posted by Tandem Affinity at 4:36 PM on December 16, 2011

If you'd dig a CSI-type slant, perhaps the loci used in CODIS?
posted by Sublimity at 4:42 PM on December 16, 2011 [1 favorite]

Great question. Pigmentation genes would be fun... or CFTR or FMR1 if you may be having babies one day - may as well make it useful!
posted by superfish at 8:32 PM on December 16, 2011

HLA was also my first thought. My second thought was something in the HARs (human accelerated regions - I think this is the first thing Tandem Affinity was referring to in his comment), but I'd imagine a lot of those could be disease candidates so that's something you'd want to check.

If you want variable but not disease-linked genes, you could probably get that by looking for regions of the genome with a Ka/Ks ratio of close to one (maybe using chimp as reference), which would indicate either neutral drift or opposing positive vs. negative selection. There's got to be a study that did this already but I can't find it right now.

Also, on the off chance you're a night owl or an extreme early bird, it might be fun to sequence CLOCK.

This project sounds awesome! Post the results!
posted by en forme de poire at 9:54 PM on December 17, 2011

Thanks for you answers so far!

FWIW, I've also had suggested to me:
Tyr Tyrosinase
Trp1, Trp2
Dopachrome tautomerase

...which I think are all involved in determining skin colour. Apparrently sequencing those four should give a result that's (near-)unique.

Due to the holidays and changing job/city, I've been quite slow on this. But hopefully I can get something sorted in the next few weeks. I'll update here if the question is still open when I have something to show.
posted by metaBugs at 4:52 AM on January 4, 2012