Behold, the Interpretome!

By Razib Khan | June 14, 2011 3:41 pm

Zack and Dr. Daniel MacArthur have already pointed to the Interpretome tool (Chrome and Firefox only!). As Daniel noted the PCA option will probably be the most fun for people. Just load your raw (unzipped) 23andMe data file, and project yourself onto different population sets. Since I know how to use EIGENSOFT this isn’t too revolutionary for me, but I’ve been putting off analyzing my chromosomes’ ancestry segment by segment. Luckily Interpretome has a much better experimental “ancestry painting” than 23andMe. Just make sure to use the HapMap3 setting to get the most juice.

I used 40 individuals to compute my results, and it took less than 10 minutes. The reference populations below are African Americans, Bantu in Kenya, Mexican Americans, Gujaratis, Tuscans, and Chinese.

CATEGORIZED UNDER: Genomics, Personal Genomics

Comments (10)

  1. Anthony

    At least some of the PCA works in Opera, too, even though they don’t advertise it. I’m not sure if HapMap somehow fails due to weirdness in Opera’s javascript, or I’m just not patient enough to let the results run.

  2. Eze

    Interesting website! I like that the PCA loads very fast.

  3. make sure to use at least 10,000 SNPs though.

  4. Bruce

    Anybody know it the PCA works in Avant browser?

  5. qohelet

    So by all genealogical and genetic accounts (23andme, Dodecad, the like), I’m fully Ashkenazi Jewish (maybe a spot or two of Sephardi) on both sides, equally. Yet Interpretome is telling me that one side is almost entirely “Tuscan” (proxy for European/Mid-Eastern, I assume), but one is significantly mixed, heavy especially with Gujarati and Mexican segments.

    How is this?

  6. pconroy


    Using HapMap 2, I was 100% Red=CEU

    Using HapMap3, and setting the threshold to 40 SNp’s, I am a mosaic??


    The problem with their approach is that they include 2 Admixed populations – African-Americans and Mexican-Americans??

    So the African-Americans hits I have are probably just Northern European – as they don’t have a Northern European reference population, for instance!

    While the Mexican-American hits I’m getting probably represent Western Europe/Atlantic Europe, and not necessarily Maya or Pima like sequences.

    Where the Gujarati probably represent a West Asian component best (ANI)??

  7. ackbark

    I just did this yesterday!

    And just now did it again tweaking some settings.

    23andMe tells me I’m 99% northern European with 1% or less African. Interpretome’s Hapmap 3 tells me I’m primarily Tuscan with scattered bits of Mexican American, Gujarati and African American.

    Hapmap 2 tells me I’m 100% CEU.

    On Ancestry by PCA I ran HGDP:European yesterday and it showed me positioned directly between an Orcadian dot and a Russian dot. Ran it today and it shows me right in the middle of the French.

    On POPRES:European it shows me surrounded by the UK.

  8. obviously some issues with reference populations 🙂

  9. Nick

    Question–and I’m new to this, so apologies if it’s a dumb question. My wife is from Central America, mostly of Spanish ancestry with some Native American as well. 23andme said she was ~80% European, ~20% Asian, plus a bit of North African (4%, I think it was). On the 23andme ancestry graph she showed up as Burushaski, which I assumed was because 23andme didn’t differentiate between Asian and Native American on the ancestry mapping graph. But on the interpretome graph, she’s showing up as Central South Asian even though we would have expected her to land somewhere between European and Native American. Can anybody illuminate me as to what’s happening?

  10. Stephen

    I had trouble using the Interpretome in Firefox. But in Chrome on the Hapmap 2, I was all CEU on half my genome and about 98 or 99% on the other half, with a couple of extremely thin YRI segments and several longer, but not long, CHB+JPN segments, which I would guess is maybe Native American from my father’s side, who descended from settlers in Massachusetts colony. On the Hapmap 2, the half of my genome that was completely CEU in Hapmap 2 was about 80% TSI with the remaining 20% GIH with a little ASW, MEX and CHD. I would guess that that half is my mother’s, who is Czech and Croatian and probably has more affinities to Central Asia than my father’s family, which is all from Western Europe. The half that had had a little YRI and CHB+JPN in the Hapmap 2 painting, was almost 100% TSI in the Hapmap 3 painting.

    What I thought was interesting in both the Hapmap 2 and the Hapmap 3, was that even though the Hapmap 3 is hard to interpret, my genome showed an extremely clear split down the middle, and that my parents ancestry can be analyzed in such a way as to make it split. Which is not totally surprising: my Balkan “cousins” on 23andme do show, even on 23andme, a higher percentage of both Southern European and Central Asian in Global Similarity than my non-Balkan “cousins.”


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar