Is it possible that ancestry.com and 23andme.com use the same lab for analysis? More to the point, how would I know if they did?
DNA sequencing has become almost a garage industry. (And DNA synthesis isn’t far behind, by the way.)
Curious why it matters to you?
Are we as a society ready to see on the Toy’R’Us shelves the Junior Gene Splicing kits (for ages 8 and above)?
IIRC, 23andme and Ancestry now use custom SNP chips (microarrays of probes that identify specific genetic variants). A quick google reveals that 23andme’s chip has 600,000, while the Ancestry chip has 700,000 probes. On top of that, they’re certainly using different analysis pipelines. There will certainly be some overlap between the two.
Beyond that, I’d WAG that each has their own microarray readers in-house. If they have enough business to keep a reader running 24/7, it’ll be much more efficient to buy the readers than send samples to an outside lab.
No, but we are right on the verge of terrorists being able to synthesize nasty viruses and the like. Want the complete map for build-your-own Ebola?
Analysis is one thing, comparison with a database of genetic information is quite another. I’ve always assumed companies like these are in the business of getting people to provide DNA samples precisely in order to build up a monetisable database.
As to why it might matter…
-
If I send samples to both, and results differ, who do I believe?
-
If results were identical, how would I know if results were accurate or was just the same lab used?
This isn’t remotely true. First of all, the notion of being able to “synthesize” an arbitary genome that can have malignant and pervasive effects on human cells is just not possible at the state of the art. All existing and anticipated bioweapon threats rely on modifying existing viruses, bacteria, or other existing pathogens. Our understanding of the mechanisms by which different viruses work is still kind of primitive, hence why it takes so long to come up with vaccines for viruses that the human immune system has not already developed a response. We can modify genomes by resequencing (e.g. using cRISPR or other more laborious methods) but the capability to fabricate a functioning genome from scratch is in its infancy.
Second, a viron is not just the genome of the virus, but also the protein coat (capsid) and lipit jacket. (There are a few known viruses without capsids that are basically just free RNA with some small lipid structures for binding and some simple polymerase, but they’re probably not very suitable as an aggressive weaponizable pathogen.) Constructing an entire virus, while less difficult than constructing the active mechanisms of a living cell, is less complicated but still well beyond current capability.
Third, in order to produce an effective bioweapon, you’d need to start with an already infectious pathogen with a suitable vector for wide contagion, enhance the virulence (but not so much it burns out before itis able to spread among geographically mobile carriers), and then cultivate suitable quantities of the pathogen for delivery, requiring a suitable host (probably primates or pigs). It’s a large scale operation to produce enough pathogen to effectively weaponize and spread that is probably beyond the means and patiences of a terrorist organization, requiring the resources of a small nation or large corporation, and there is the ever-present risk that technicians overseeing the production may become infected themselves, especially if they are not experienced in working with infectious pathogens in a production lab environment.
None of this is to say that this capability won’t be available in the foreseeable future to sufficiently capable parties with adequate resources, but the notion of building designer Kil-O-Virus pathogens by small groups of terrorists is just not a reality. Nor do they need to; there are plenty of pathogens, some residing in the dirt just below your feet, or in non-treated water supplies, that could potentially be weaponized and used as a terror weapon by a knowledgeable and devoted organization. That so few have is an indication of the lack of technical expertise and resources for these groups to actually follow through on bioterror weapon development.
Stranger
This really isn’t accurate. Viral genomes are small enough that chemical synthesis isn’t a significant challenge anymore. To take one example, the ebola genome is only 19 kbps. Direct chemical synthesis isn’t quite good enough to make the whole thing in one go, but it’s easy to synthesize a half dozen smaller fragments and assemble them into one piece.
With the resources of the small molecular biology lab that I have at my disposal, I could buy chemically synthesized fragments of the ebola genome for $3500 from IDT, assemble them in one hour in a single tube, and transfect the assembled genome into mammalian cells. Within two weeks I’d have cells that were manufacturing intact ebola virus. Most of that time would be waiting for the delivery from IDT (or, more likely, a visit from some Very Serious Men With Guns).
This approach may not count as “synthesis” since it uses living cells to bootstrap the virus production, and not just organic chemistry. However, that’s not really a useful distinction, in terms of countering bio-terrorism.
Thankfully, there are safeguards. IDT won’t sell you such clearly dangerous DNA sequences without a shit ton of red tape, oversight, and approval. But IDT doesn’t have magic technology. They use fairly standard synthesizers that are merely well-automated and streamlined. With more time and labor, similar results could be achieved with off-the-shelf DNA synthesizers. It’d take only a handful of PhD molecular biologists and a perhaps a few million dollars to set up the lab.
Given how easy it is to manufacture a virus, I suspect that we have counter-terrorism efforts to thank for the lack of any significant bio-terrorism attack.
I should probably stop contributing to the bioterrorism hijack (heh) and address the OP’s questions.
What do you want this information for, anyways? If it’s for personal edification, I’d start by believing the results whenever both tests are in agreement. When there’s a conflict, you can dig into the raw data to get an idea about whether one of the tests was in error. E.g., one company says you have Variant A, but the other company says you have Variant B. Some of these conflicts are inevitable due to errors in the testing methodology.
If it’s for something more consequential, like a substantial disease risk, go to a doctor to get tested properly. Then, talk to a genetic counselor to understand how your particular genetics will affect your daily life.
Sure, but what you are talking about would require an extensive lab and production facilities to actually produce and weaponize a useful amount of pathogen, and that could just as easily be done at that point with readily available naturally occurring pathogens, e.g. botulinum, e. Coli, et cetera. The idea of being able to tool up a tailor-made supervirus capable of defeating all natural and health organization safeguards is, at this point, still mostly science fiction. It may not remain that way as our abilities in genetic engineering and understanding of virology and molecular biology become more sophisticated, but for now the terrorists are probably going to stick to throwing bombs or dispersing radioactive materials to produce the desired effect.
Stranger
Well, some pathogens are more dangerous than others, and some of the most dangerous, we’ve put enough of a priority on them that they’re no longer easily available naturally. Consider smallpox, for instance: If a terrorist organization wanted smallpox to weaponize, they could break into the highly-secure facilities in the US or Russia to steal actual samples… or they could get a copy of the smallpox genome over any information channel, and re-synthesize it. Which would be easier for them, and is that a problem?
Again, there is more to a virus than just the genome. For an Orthopoxvirus like variola, the genome is enclosed in a large capsid that serves to attach to the glycosaminoglycans of the host cell and mediate cellular endocytosis (transfer of core viral material into cytoplasm). Even if you could reproduce the DNA polymerase to get a host cell to produce new viral DNA and protein material, there are other regulatory mechanisms to force the cell to preferentially reproduce the viral DNA. This is why vaccines use “deactivated” viruses or similar but benign viruses to stimulate human immune response.
Viruses, like all things at the basic mechanical level of cellular and molecular biology, are incredibly complex systems of vastly interrelated functions, and we just don’t have the skill or even basic knowledge to truly just synthesize one from raw organic materials (yet). And frankly, smallpox, while virulent, has a significant evolved immune resistance in many human populations who have engaged in animal husbandry. The mortality rate is not that high and rate of natural mutation is low, so unless the virus were radically altered it would not be an effective basis for a bioweapon (hence why Soviet research on smallpox as a weapon lapsed in favor of other, less well developed pathogens).
Stranger
The capsid isn’t necessary to get the viral genome into the cell. Cell biologists have a whole host of techniques to force any DNA sequence into any cell. These are trivial, decades-old techniques.
All of these mechanisms are encoded by the viral DNA. Once the genome is in the cells, the host will transcribe and translate all the viral proteins.
If the virus requires some of its proteins for the initial transcription, translation, etc., it’s *trivial *to provide those proteins, using DNA constructs that are reliably expressed by the host.
I don’t use the term “trivial” lightly, either. Getting a cell to express an arbitrary set of proteins is the sort of thing I assign to undergrads as their first introduction to real lab work.
I’ll defer to your applied knowledge, but I still don’t think that creating a pathogen of novel construction or highly enhanced virulence is practically within the ability of terrorist organizations, and the weaponization of biological agents in general has been less effective than the effort has warranted compared to other strategic weapons. A terror organization, of course, has other goals (to create fear and unrest), but there are naturally occurring, readily available pathogens which could be used with essentially no modification for this purpose.
Stranger
To be clear, I’m not worried about some genetically-engineered super-smallpox with a custom hand-written genome. I’m worried about plain old ordinary smallpox, like what used to kill millions before it was eradicated in the wild.
And yes, the Soviets rejected its use as a bioweapon, but a superpower nation like the USSR has very different requirements for a bioweapon than does a non-national terrorist organization. Smallpox would still make a fine terror weapon.
The biggest stumbling block is still the synthesis of the pathogen’s genome. The smallpox genome is about 186 kilobases long. Even if you could convince a synthesis company to produce complete smallpox, it would cost hundreds of thousands of dollars and is not even an efficient way to do it. Constructs of that size need to be synthesized as short fragments and assembled in vitro. It’s absolutely not a matter of simply keying the desired sequence into the synthesizer. Also, as lazybratsche indicated, the various DNA synthesis companies screen customer sequences to ensure their product is not being used for harm.
Commercial DNA synthesis is a highly specialized chemical process and many of the details that enable long-length synthesis are trade secrets. It’s not something that you can set up as a turnkey system. A hypothetical bioterrorist would need to reproduce all of the current R&D before they could even begin synthesizing a bioweapon.
My sister and I (and a couple of others) used 23andme. I have no reason to think it was anything other than accurate.
The main reason for the testing was curiosity about ancestry. Like all American mutts, the story on both sides of the family is that we have native American blood. However, in all the family tree research that has been done, there isn’t anywhere you can point to and say ‘there it is’. The ancestor results showed the same thing, no native American. I understand that a single ethnic ‘outsider’ can be ‘washed out’ of your DNA after maybe 6 or 8 generations, so maybe I’m descended form a Cherokee Princess, but it was way back. I wouldn’t bet any money on it.
It pegged my sister and me being siblings, despite different last names, and some separation in process dates.
23andme also has ‘DNA Relatives’ - you they tell you who else in their database you are related to, and how closely. I’ve made contact with a couple of 3rd and 4th cousins I didn’t know I had. DNA Relatives is an opt-in feature.
The main reason I went with 23andme over ancestry is that ancestry is owned by the Mormons, and they don’t believe that us humans have Neanderthal DNA. They call it something like ‘undefined’ or ‘other’. 23andme will tell you how much you have. I have 279 Neanderthal variants.
23andme also has FDA approved health screening info, but I’m not sure it’s worth the extra cost. Ancestry info is $99, ancestry and health is $199. But the health info they give is for some mostly uncommon syndromes that if you have them, you probably already know. The big disease marker that most people want to know is BRCA, and they don’t test for that (I understand that the test is copyright).