Tag Archives: 23andme


Friday SNPpets

This week, DNA was indicted and decades later led to a conviction. Genomes of birch trees and shape-shifting butterflies. And the most interesting stuff to me is non-human, but dbSNP will stop accepting non-human species info. Sigh. Well, I do think alternative splicing is interesting too, and we have some of that this week. Human gene editing, tumors, and various types of personalized genomics were out and about this week as well. It’s not that I dislike humans (all the time), but I’m sort of agnostic on species.

SNPpets_2Welcome to our Friday feature link collection: SNPpets. During the week we come across a lot of links and reads that we think are interesting, but don’t make it to a blog post. Here they are for your enjoyment…

Friday SNPpets

This week was a busy and diverse week. From Bioinformatics as an Amusement Park, to the very serious FDA approval of 23andMe’s reports. From Lenski’s April Fool’s post to cancer databases. From diagnosing children with genomics details, to CRISPR sci-fi. We live in interesting times. But at least now we will be able to see everybody citing everybody else.

SNPpets_2Welcome to our Friday feature link collection: SNPpets. During the week we come across a lot of links and reads that we think are interesting, but don’t make it to a blog post. Here they are for your enjoyment…


Friday SNPpets

A pretty typical week, lots of great software and fascinating data. Methylation analysis, 7-set Venn, a new genomics comic strip, some personal genomics, some clinical genomics.  Gosh, I love my life. But the most amusing and unexpected snippet: misuse of a Circos image of caught the eye of the developer, and he found that Jurassic Park dinosaurs are made of corn.

Welcome to our Friday feature link collection: SNPpets. During the week we come across a lot of links and reads that we think are interesting, but don’t make it to a blog post. Here they are for your enjoyment…


What’s The Answer? (23andMe to other formats)

Biostars is a site for asking, answering and discussing bioinformatics questions and issues. We are members of the Biostars_logo community and find it very useful. Often questions and answers arise at Biostars that are germane to our readers (end users of genomics resources). Every Thursday we will be highlighting one of those items or discussions here in this thread. You can ask questions in this thread, or you can always join in at Biostars.

This week’s highlighted question was from someone with personal genomics data in their hands, but doesn’t know what to do next.

Question: How to I convert 23andMe Raw Genome to GenBank or FASTA?

I used 23andMe to download my raw genome. I have it in a .txt file but you can’t use the format for real bio programs. i want to make my own library for further analysis. Does anyone know how i can convert .TXT to FASTA, GenBank, or any other usable file type?


Although Biostar usually hosts questions from folks who are a bit more advanced in their grasp of file formats, this question struck me as interesting for a couple of reasons. The needs of folks who are not practitioners, but who find themselves with data in their hands, will only increase going forward. And although the companies will offer some tools, there’s a niche for some lighter-weight public tools. I discussed this before on the issue of genome browsers, which are currently too much heavy lifting for intro-level users. I know there are some open data communities forming around this data too, but so far what I’ve seen has been more sophisticated early adopter types.

But I imagine it would be difficult to get funding for such intro-level tools. They probably wouldn’t score well on “innovation” and some of the other traditional grant criteria because–well, because that’s not what the system does.

Maybe it would make some good class projects for some coders who are learning to build tools, and to work with this type of data. Make some gentle 23andme to X-format converters. A browser that’s not too hard to load your data up and look around without too many tracks. These folks are going to need more hand-holding. They don’t know what formats they need, or what is available for them to do.

But have a look at the answers, and if you have other guidance for this newbie, drop some comments over there.

What’s The Answer? (23andMe bioinformatics)

BioStar is a site for asking, answering and discussing bioinformatics questions and issues. We are members of the community and find it very useful. Often questions and answers arise at BioStar that are germane to our readers (end users of genomics resources). Every Thursday we will be highlighting one of those items or discussions here in this thread. You can ask questions in this thread, or you can always join in at BioStar.

This week’s highlighted Biostar item is an incarnation of the 23andMe drama. I’ve seen people take on the legal, medical practice, and social aspects. But this thread takes on the underlying bioinformatics as an issue. I thought that was interesting.

Forum: FDA sends a warning letter to 23andMe – Personal Genomics Service marketing to be discontinued

There has been an interesting development in the world of personal genomics. The FDA sent what may be interpreted a cease and desists order:

Therefore, 23andMe must immediately discontinue marketing the Personal Genome Service (PGS) until such time as it receives FDA marketing authorization for the device.

Read more here: http://www.fda.gov/ICECI/EnforcementActions/WarningLetters/2013/ucm376296.htm

This seems to put bioinformaticians at the forefront of deciding what constitutes credible evidence that the PGS does indeed work as advertised.

Istvan Albert

I added my favorite additional pieces over there, which haven’t got much airplay among the genoscenti. But go have a look at the wider discussion among bioinformatics practitioners, which I thought was interesting.

SNPTips update (1.1)

I did a tip of the week on SNPTips a few months ago (more information there). It’s a great addon to view your genomic data while browsing databases and web sites. They’ve moved to version 1.1. There are two nice new features and some bug fixes. The features are:
*You can now use your deCODEme data, in addition to the 23andme support they started with.
*You can use SNPTips even without raw data to view SNPs on a page.
*and it’s been updated for Firefox 4.x.

You can check our our previous tip here (which still applies :).

SNPTips landing page at 5am Solutions.

The cost of genomics, revisited

1,000 dollars to sequence the entire 3 billion ‘basepair’ long human genome is the holy grail. Last year, on this blog, I predicted that we’d reach that goal ‘within a year’. Well, it’s been a year and we aren’t there. I did hedge a bit in the comments later, but the main point was that the cost was plummeting and headed for $1,000 for an entire genome. The graph at the left is cost of a single genome from 2003 to 2010… 300 million, to 10 thousand.

The cost is still plummeting. Illumina* has lowered the cost of sequencing the genome to $5,000. Half.Complete Genomics* today announced that their first quarter revenue was 6.8 million, up from 300 thousand last year. They have 2,000 complete genomes to sequence in their backlog. They charge < 10,000 for small orders, down to $5,000 for bulk orders (in research groups). The cost of sequencing large groups for research is getting much smaller.

So the price plummets still.

The price for genomic scans also plummets. 23andMe, which scans about 1 million variable sites in your genome (most of the human genome is not variable from person to person), cost 400 dollars last year, in December they price went to 200. After a one-day free sale, it’s now $100.

Of course, the question remains how expensive will it be to analyze that data.


*full disclosure, I own a (very) small amount of stock in both Illumina and Complete Genomics.


Gene Envy. I haz it.

Last night I had a fun conversation with some non-scientists on personal genomics. A lot of my conversations on this topic are “inside baseball” with people who know the field (and the minefields) really well. So it’s fun for me to explore the larger social and ethical issues with people who are just beginning to think about them. And they coined a phrase to describe my feelings on something I had seen: gene envy. They were right :)

We talked at some length about what you can and can’t learn from the “reader’s digest” version of your genome that you get from a SNP scan like 23andMe provides. How SNPs can be informative, but are not the only types of variations that might be important in your genome, and they certainly don’t represent all the SNPs possible. We talked about the kinds of things you can learn from the SNPs right now–silly trait sorts of things like whether you have wet or dry earwax (as if you didn’t know already), as well as some information about potential disease incidence and response to certain medications–but even that data is sometimes preliminary and subject to change still.

That’s all the stuff I expected. One thing I wondered about was if I was going to be affected by knowing about other people’s genes…and how I might react to that. Well, now I know. I haz gene envy. [for my color-blind siblings, I wrote that in green, btw]

I hate to admit it. I want to be better than that. But maybe I lack a gene for that or something…. I should state right off that I’m not mad at anyone personally–nobody has control over what’s in their genes. It’s not about them per se, it’s about my reactions.

I’m not a hugely introspective person, but I have been trying to pay special attention to my responses to the new landscape of personal genomics. We have a front-row seat to this, and a decent copy of the program. So I feel a responsibility to be aware and honest about what my reactions are. And this is one I don’t like in myself. But it’s there.

I’ve downloaded some other people’s genome scans that they made publicly available. I use them to test out some of the software that I run. And if I want to make screen shots of stuff I refuse to reveal my SNPs right now as I don’t feel safe doing so.

One day I was looking around at obesity variations for some project I was working on. I had forgotten that I had my SNPTips tool on. When you have SNPTips on and search the medical literature at PubMed, the abstracts can be a bit of a roller-coaster. You can read about this research, and roll over the SNP IDs and be shown your own variations (or in my case other people’s). It looks like this as you are reading:

Note: that is NOT the paper I was reading nor the SNP and gene I’m referring to; this was just an example image.

So I am reading along the text, and check the genotype of this individual’s SNP. In that case, a result of AA meant that you had the “skinny” version of the gene in that study. It bugged me. I rolled my eyes, and thought of course it’s easy for this person to be thin….sigh. I was jealous.

In another case I was relieved that someone’s genes didn’t carry a variation I have. I saw the results for someone and thought: oh, good, they probably won’t have peanut allergy like I do. I’m glad for them.

One time when reading in the 23andMe discussion boards I found myself looking at the results for people who are probably resistant to HIV. I thought: nice for them. But the ones who didn’t have that–I was sort of shocked to see them telling everyone. I wonder if that’s wise when we know that data scraping is going on at some medical social networking sites. I easily could have written down their results and their names if I had been an insurance investigator. (If you want to check your own data there, search for CCR5 at 23andMe.)

Another “variation” on my response came when the Alzheimer’s alleles were recently released by 23andMe. I saw several people celebrate their alleles. WOOT!! Even before I looked at mine I resented this. Yeah, I think, you lucky dogs: goody for you. Now shut up and stop rubbing it in. Some people are not going to have a good outcome of this.

And I think to myself: it’s easy to be cavalier about the results you get from personal genome scans if you are healthy and have good genes–and stuff that won’t affect your insurability. Or if you live in a place where your health insurance is guaranteed or you are otherwise secure in your insurance. That’s not the case for all of us.

What I’m feeling is a mixture of things. I’m jealous and irritated by other’s “good” genes. I’m relieved in some cases. I’m anxious for my own outcomes in others. Even though I know that what we know today is only part of the story in most cases, and even with “bad” genes there are probably other modifiers or interactions that may mean you can get off scot-free with your alleles. But will insurers want to take a chance on me? I am skeptical that they’ll lean to the liberal side of the interpretations.

I still think we’re likely to see what I called Genorexia Nervosa–where some fraction of people get so wound up about their genes–or their kids’ genes–that they go overboard with how they think they should eat, or exercise, or allow themselves to be exposed to. (There’s probably an OCD sort of gene that influences that….). But I didn’t see the personal envy coming. That’s what happens when you have a front-row seat on a roller-coaster. And that’s exactly why I ended up doing the 23andMe scan–to join the ride, for good and for bad.

I’ll get over it. Although I do think I have a variant of the hold-a-grudge gene. I may mutter things about your skinny genes if I see you in skinny jeans….

Your genome scan for Free*

*while supplies last, plus 9 dollars a month for one year.

23andme is having a sale (again, I did our family’s scan when it went down to 100 late last year). This time the up front cost is $0. Of course you have to pay for shipping and the 9/month fee for updates (for one year). Still, that’s nearly a 2/3rds price reduction from just yesterday when it was $200 plus the 9/month fee.

They are doing this in advance of DNA Day, a national day to raise awareness among students and the public about genetics and genomics. The sale is today only (or while supplies last :).

So, if you had been considering a dna test, maybe now is the time to do it. And then come back to Genomes Are Us and we’ll help you decipher it :D.

This got me to thinking why 23andme is cutting their prices so drastically so often. I’m sure part of it is a market share and marketing ploy. I’m wondering also if there isn’t some thought that the more average citizens invested in the technology, the more likely they’ll fight draconian (in their eyes) regulation. But I’ll bet it’s the former more than the later, but it couldn’t hurt to have more people invested and interested.

Well, I did my duty and ordered my maximum of 5 tests :D.



(crossposted @ Genomes Are US)


Tip of the Week: SNPTips and viewing personal genome data

Today’s tip of the week is on SNPTips. We had a guest post on this earlier. We usually do tips on databases and analysis tools, but after getting our 23andme data, we’ve been using SNPTips often and thought it might be of use to some of our readers. SNPTips was created by 5am Solutions for 23andme* customers to easily view their genomic data while browsing the web. The tip will quickly show you how to install the browser extension and what it does. At the end of the tip, I  briefly show a custom annotation track I created of my 23andme data using UCSC Genome Browsers** Personal Genome SNP format.  The format is not perfect for 23andme data (doesn’t allow for rsID field, has fields of little use with 23andme data, etc), but it does help tremendously if you want to browse your data with the genome browser. You basically take the 23andme data that looks like this:

and rearrange the columns, add a few to make it look like this:

You can do this in a spreadsheet program like I did, it’s a bit labor intensive. If I decide to do it for my daughter’s and husband’s genome data (which is a distinct possibility), I’d created a perl script to change the format (or maybe there is something already out there?).

It basically entails:
*eliminating the  rsID column
*rearranging the columns to the correct order
*adding “chr” to the chromosome number
*adding four columns, 1 with the number of alleles, 2 with 0′s (frequency data the 23andme data doesn’t have)
*changing the genotypes from xx to x and xy to x/y.

Remember also that the 23andme position data is from build 36 (2006, hg18) and the genotypes displayed in 23andme data are oriented with respect to the positive strand on the reference assembly.

It’s not the most elegant solution, but it works and nicely with SNPTips. It has been quite addictive for me :). I’m sure there are more elegant ones that can be done.

*OpenHelix and it’s employees have no commercial connection or financial interest with 5amsolutions or 23andme.
**UCSC sponsors tutorials and outreach with OpenHelix through as a subgrantee.