Whole genome sequencing for $700-750?
Friends, I'm going to try to negotiate a deal with one of the companies offering whole genome sequencing to get the price down well below $1000.
How many of you would be willing to pay around 700–750 USD for whole genome sequencing, without interpretation?
Thanks,
Brian
-
Thanks for your work on this Brian.
I would be interested at that price point. However, it may be even more interesting if I knew of some good resources for interpretation.
Do you know of any?
___
Maximus -
The ones I've so far looked at don't seem to be very good. Still on the "to investigate" list:
- BioViva's new service (not quite up and running yet, but will be soon).
- Health Nucleus. (They offer much more, however, which means one pays several thousand dollars for services one might not need.)
-
I signed up with Veritas Genetics (George Church backed I believe) in the early days for $1k. They got so swamped that my genome was sequenced but interpretation took a year or two. Upshot was they ended up giving it to me for free 🙂
They're good, I got a long phone call with a genetics counselor which was useful, and the online reporting which is good. I've also been a early days 23andme customer (with the 1st gen chip which has a bit more testing). They both agree, and the only thing is that Veritas reported one extra potential 'issue' that was not tested for by 23andme. So for my 23andme was just as good. Now my report is very clean - I picked good parents - so if I had more issues the whole genome might be more useful.
The other thing to realize is that genome testing isn't a binary switch - e.g. Veritas only tests for single errors. For example, if you have a long swath of genes that are altered that is not tested for. This is true for all whole genome tests, the point being the Genetics counselor recommended that if there are specific concerns then you should get a specific test targeted at those genes, because that will give the most accurate results, not the generic partial (23andme) or whole (Veritas) testing. Also, you have to be careful. For example the apoe4 variant (there are several versions) is an infamous defect for contributing to Althzhimers in the worst version. But, you can mitigate the effects with lifestyle (diet and fasting) and live a long healthy life. Furthermore, a few years ago, it was discovered that this variant only causes the effect with females, not males. Long story short, be prepared to do research and dig into the results, and even then be left with "well, it's my life and lifestyle after all".)
FWIW I found genetics testing marginally useful. One problem is that gene expression plays such a big part in this - that is how you live your life determines how your genes express. So just because you have defect X - well, what does that mean? It means you miscode for (most likely) some enzyme used for something, like liver or lung function. So you get a partially effective enzyme - it's a bit of bad programming, but usually partially defective . But how that means for you is up for grabs. If you live a healthy lifestyle, then generally miscoded enzymes are compensated for. And we don't generally know what effects each defect (4 variants possible) result in.
Maximus, on interpretation, there are a lot of free and paid options. In many cases there are specialized ones, like for MTHFR variants. And you can also use the python tool that the researchers use (it's fun to see how you're related to squids). Best is to get your report, find anything concerning, then dig into that.
-
Dan Mc said:
The other thing to realize is that genome testing isn't a binary switch - e.g. Veritas only tests for single errors. For example, if you have a long swath of genes that are altered that is not tested for.Dan, thanks for the interesting reply. But I'm not sure I understand the above. Whole genome sequencing gives you a read of your entire genome, no? It's not really looking for "errors" – or am I misunderstanding?
-
BrianMDelaney Well to be clear the more I dig into this the more complicated it gets, and I only have a surface understanding. But with that no, basically while it does give you what looks like a printout of you, with nice clean answers the reality is (AFAIK/Understand) that has some degree of accuracy - and that accuracy can be variously categorized. So say you look at Chromosome XYZ/allele J - what you see may not be what it actually is. I've seen discussions of this online by grad students and such, there are different ways of doing the test, all with different kinds of accuracy. For example, the Veritas councilor was very clear that they only test for single SNP errors ('snips' single nucleotide polymorphism) - think of this as 1 bit errors if you're a CS type. This doesn't test for say, an entire region that is messed up (maybe a prion snipped out a long section when you were a zygote and substituted something else). You might get a slew of gene 'defects' - SNP variants that is, but it's actually because of a whole section problem, not just what looks like a few random errors. Everybody has tons and tons of 'defect's', or better 'variants', four variants for each SNP.
Practical example: say you're a woman and you come up with a bad variant (SNP) on a gene for breast cancer. Most defects are 'maybe a problem', but this one (I don't know the name of offhand) is 'really a problem'. Some women see that, freak out and have their breasts chopped off. A better response is to take another genome test, one that is targeted specifically at that defect. A defect, isn't a defect, isn't a defect. The SNP you are seeing could be from a single bit error, or from a region error, or something else. Now you're getting at the question 'yeah sure, but is what I see REALLY my genome, or not?'. Truthfully I'm not entirely sure the answer to that - I think yes, maybe? In practice the rule is 'take the warnings from the report, if any are concerning considering lifestyle, say you smoke and have a genetic disposition to lung cancer, then do a focused test to get a better answer'.
tl/dr
But even more complicated, genes work in concert. Say you are a MTHFR, turns out it's not one gene, it's something like 20-30 that play in this. So really you need to look at all of them to get an idea. And at best that's what you get - an idea. The gene expression is a complete unknown. One, because it changes on even a daily basis. For example, when you do a fast there is a set of genes that get activated only during that time, and are otherwise dormant (they are genes involved in protective cellular enzymes).
Anyway, hope that helps. If nothing else do 23andme which probably gives you 90% of what you want. If you're a 'no holds barred do everything possible' person like you and me, then do the whole genome. If nothing else it's a nice security blanket. For me it only gave me a single extra possible red flag. If you're somebody like Kurzweil with loads of money to throw at this, I'd do every whole genome test publicly available, and also all the focused tests, and I'd have a team to analyze it.
-
-
Dan, thanks. I'm not worried about the interpretation, just about the accuracy of the read, which does indeed seem complicated. I'll have to look into this much more deeply. I'd always assumed, for example, that the problem of relating the definition of a given spot on a gene to a given spot on a chromosome (distance from one end), via RS numbers, had been solved in some way, which is why we have the GRCh assemblies. But maybe not! As you note, if there are deletions (or insertions), the position from the end of the chromosome will be changed, and, if the deletion is within a gene, the position within the gene will be altered.
I will look into this more fully.
Meanwhile, I just found a $600 whole genome deal:
https://www.bgi.com/global/human-whole-genome-sequencing-600/
But not sure about the quality of the sequencing. I will contact BGI for details within a week or two, and report back.
-
BrianMDelaney yes let us know what you learn. I may be getting it all wrong and am going from my limited information (I'm not familiar with the terms you use above for example), but what was clear in my discussions with the genetics counselor and my minimal research was that accuracy wasn't guaranteed, or at least you had to understand what accuracy was guaranteed.
-
-
Dan Mc, please verify the comment made above:
"Furthermore, a few years ago, it was discovered that this variant (apoe4) only causes the effect (increased Alzheimer's risk) with females, not males."
Could you provide the source of that finding as I've never read anything like it.
Isn't the ap0e4 variant also involved in increased risk of cardiovascular disease? Is that risk limited to females as well?
-
Ozone8 A search is your friend, keywords "apoe4, Stanford, male" etc, at the time it was reported in a Stanford internal magazine (on the web) and elsewhere in a paper. In 2016 or so a researcher reran through the old data and found what I said above, my genetics counselor knew about it and also discounted any concern.
Haven't heard about a cardio link. At any rate like I say it's all epigenetics anyhow - best approach is to live a healthy lifestyle, test and verify, and use your report for focus areas.
-
-
Apolipoprotein E4 (produced by the gene variant APOE-ε4) was, originally, going back several decades, known as a factor that increased risk of cardiovascular disease. Only later did Mahley discover its role in Alzheimer's and (a bit later) dementia in general.
In any event, male APOE-ε4 carriers appear less at risk for dementia than female carriers (not sure about cardiovascular disease), but the balance of studies indicate they have a greater risk than non-carriers. (And APOE-ε4 homozygotes, men or women, have a much greater risk.)
But Dan is right that lifestyle factors can overwhelm most genetic risk factors, even APOE-ε4.
-
I just found a great resouce:
I won't have time in the near future to go through the (huge!) list of sequencing providers. If someone else can take a look that would be appreciated. Otherwise I'll try to get to this in December.
-
For anybody following this, Veritas Genetics, where I got my whole genome done, is running a sale for $199, first 1000 sign ups. Run quick if you want a great price. They're excellent, George Church backed. Go quick if you want a cheap whole genome!
-
BTW one benefit of a genome report like I got at Veritas - see if your provider does a drug report (Veritas does this). Since us anti-aging super-advocates are venturing into pharmaceuticals it can be useful. For example I've reported elsewhere here to check your CYP3A5 gene - if you have the CC variant for example you may have decreased metabolism of Rapamycin, so you should consider a lower dose.
-
Evidently it is too late for the $199 special as the price on the website is showing $999 now.
-
Ozone8 Wow, that went fast, sorry!
-
-
By the way for those who have signed up, with Veritas (and probably others) you can pay a small fee and get the raw XML data for your genome, in addition to the free formatted reports. Just request them when your genome is processed. The data come via a sftp connection of about 500MB.
For viewing you can use the free IGV (Interactive Genome Viewer) Java application which is available on Mac, Linux and Windows. There are probably other tools and services you can use with this data, this one is for use by researchers and is quite interesting. You can easily troll through your genome and find out which genes you share with squids, for example.
- To use the tool load your data
- It creates an index (takes a few minutes depending on your system)
- Human hg19 is selected in upper left hand
- The next drop down selects chromosomes (AFAIK)
- You can zoom in and out upper right hand corner
- On the "RefSeq Genes" display on the bottom, click on the symbol (e.g. PATJ LIN00466 ...) and it will jump to the nih web page on that with all relevant data
Further I'll mention again that my understanding is that all tests have limitations. For example, with Veritas (and probably others) it only tests for single errors not sequence errors. For example, say the canonical sequence for some part is "A D C A D D T T", but for you that sequence happens to be completely knocked out, and you happen to have "A D D T T T C C" which happens to match on the first two positions. Veritas will report that as having two good with the others being defects, and does not detect that the whole sequence is bum, and you randomly got some matchups.
That's my understanding, if you learn more about this please share your information.
-
BrianMDelaney OK I'm walking back my assertion that genetic testing has marginal benefit. After hearing that Bill recommends fasting Glucose lower than 85 I've been testing mine (home kit) and it has been coming in 85-100, up to 110 if I ate a starchy dinner. My lifestyle is ideal - exercise every day including cycling/running/pilates and a whole low calorie foods vegan diet, and rest of my markers are also ideal, so why isn't the glucose even better (note that you're not officially considered prediabetic until > 100, which I can easily hit). So looked in my full genome report and found three genes all which predispose to elevated fasting Glucose. Combining with other soft measures (e.g. lifelong dietary preferences and responses to foods) I believe that these three genes drive up my glucose, which is partially managed by lifestyle.
So I think it's probably safe to say it's not necessarily a single gene but combinations of genes that play into macro aspects of you. For example, on telomere length there are two that say longer, two that say shorter, and one that says no difference. Assuming equal effect then it might be safe to hypothesize that in this example average telomere length is most probable.
-
DanMcL said:
so why isn't the glucose even better ...Perhaps you would benefit from a particularly low carbohydrate diet. Replace some carbs with plant-based fats.
-
cacarr Thanks for the suggestion but I tried that, makes no difference. Glucose management overall is excellent as is A1C and I take back what I said above about 110 and a starchy dinner, that was just a fluctuation. I have a problem with 'morning phenomenon' which is a 3AM glucose release from the liver; mine appears to be doing it too much, and I think these genes are the culprit. Hoping that Metformin can control that, today I go from 500mg->1000mg.
-
-
FYI - you can now get you whole genome sequenced for FREE here:
It associated with George Church and Veritas Genetics.
Also download your data (from any of the sequencing sites) and upload here to get an amazingly detailed view of all the identified issues in your genome. It only cost $12.
BTW I'm not affiliated with any of these but I thought you folks might find them useful :)
-
Ul Looks like its $99 now, and $199 normally, which is a great price. Looks like they want to really start finding out what the genome does, but I think they'd need detailed medical history to be able to do that.
-
DanMcL There are two buttons on the https://www.nebula.org/ page. One is "$99 buy now" the other is "Get started". If you click the "Get Started" it's free.
-
-
Isn't that for a very low pass WGS?
I completed all available surveys on the site and only earned 550 points. 1000 points (or $99) are needed for the free test. The only way to earn enough points appears to be getting 4 or 5 others to participate in the Nebula program.
-
https://www.dantelabs.com/collections/our-tests/products/whole-genome-sequencing
I've done it last year :)
-
Am I the only one waiting a LONG time to get results from Veritas? I'm on day 73 of the "analysis" phase.
