PDA

View Full Version : Epigenetic testing for the masses (New Scientist)



MacUalraig
02-20-2019, 02:04 PM
There was a brief article on this industry in last week's New Scientist and two of their journos are trying out a couple of companies, Chronomics and Muhdo. They are getting their tests paid for.

Chronomics appear to offer their test with or without a WGS behind it. It was formed by a bunch of Cambridge graduates although when I checked their company accounts it seems a lot of them later resigned as directors. To get the WGS bundle it appears you would have to purchase the 'Genome Rockstar' product:

£699.00 / year
+ £799.00
/ one off whole genome test

https://www.chronomics.com/products/pricing

Company records filed so far:

https://beta.companieshouse.gov.uk/company/11120038

In the interests of balance the article quotes a guy from the US CDC saying 'no validation and no utility as far as I know'.

I am interested though to see whether they use a UK lab for their WGS as that might be appealing as a DTC option.

MacUalraig
02-20-2019, 05:24 PM
I checked with the firm and their WGS test is conducted on 'mainland Europe'.

ianz91
02-20-2019, 08:02 PM
They have ancestry info but only in the rockstar genome version of the kit lol

MacUalraig
02-21-2019, 03:45 PM
I did find this page on Illumina epigenetics options and under the array version it mentions 850k CpGs covered which might be what Chronomics are hinting at when they mention '1000000 positions covered' although its hard to make out which option they are talking about.

https://www.illumina.com/techniques/popular-applications/epigenetics/dna-methylation-analysis.html

Ysearcher
05-29-2019, 07:28 PM
Pat, here, aka Ysearcher on other forums. I have come across Chronomics recently. They don't offer much on their web site to inspire confidence that their product has been validated and peer reviewed - all I found were a few links to press reports, zero links to research validating their technology. I think I have more confidence in the Illumina product, but no idea where to find that as a consumer service.

Thanks for posting this. @MacUalraig, am I reading correctly that you have had the Dante Labs Nanopore product? Were the results useful, intelligible, what you expected???

MacUalraig
05-29-2019, 08:15 PM
I have the Chronomics test ongoing, got a message from them the other day to say sorry for the inevitable delays. I have got the nanopore in fastq format and I aligned it but I haven't had time to do any variant calling as my Chromium linked reads test arrived virtually the same day so I'm struggling to keep up with all the work I need to do!

xenus
05-30-2019, 01:16 AM
I have the Chronomics test ongoing, got a message from them the other day to say sorry for the inevitable delays. I have got the nanopore in fastq format and I aligned it but I haven't had time to do any variant calling as my Chromium linked reads test arrived virtually the same day so I'm struggling to keep up with all the work I need to do!

I know that it's not the most expensive hobby around but is there a reason you've done so many tests recently? I can see jumping on new generations of tests but just from your forum signature and this post you've gotten a ton of tests this year. I'm just curious and you can disregard this if it's something personal.

MacUalraig
05-30-2019, 07:26 AM
I know that it's not the most expensive hobby around but is there a reason you've done so many tests recently? I can see jumping on new generations of tests but just from your forum signature and this post you've gotten a ton of tests this year. I'm just curious and you can disregard this if it's something personal.

More technical curiosity than anything specific eg a particular disease.

DMXX
05-30-2019, 09:35 AM
Consumer testing of the epigenome will become a very lucrative market moving forwards.

Unlike your genome, the epigenome (the series of chemical signals that modify gene activity either directly or indirectly) is subject to change within the individual's lifetime. If I went from eating low-grade McDonald's cheeseburgers daily for a decade and then began daily meditation, exercise and a high vegetable diet, that'll reflect in the epigenome.

I can't see how epigenome testing would be useful for ancestry purposes (too many of those methylation-demethylation sites are going to be different between me and grandma), but it'll be a primary focus for health-conscious testers.

It's worth bearing this in mind, should the design of those tests be "gamed", or the advertisement to favour frequent testing, for commercial purposes (the same way older lightbulbs weren't 100% Argon - the main inert gas used IIRC - to prevent them lasting forever and ensuring a reliant customer base formed).

xenus
05-30-2019, 10:24 AM
More technical curiosity than anything specific eg a particular disease.

It is interesting from that perspective.

xenus
05-30-2019, 10:25 AM
Consumer testing of the epigenome will become a very lucrative market moving forwards.

Unlike your genome, the epigenome (the series of chemical signals that modify gene activity either directly or indirectly) is subject to change within the individual's lifetime. If I went from eating low-grade McDonald's cheeseburgers daily for a decade and then began daily meditation, exercise and a high vegetable diet, that'll reflect in the epigenome.

I can't see how epigenome testing would be useful for ancestry purposes (too many of those methylation-demethylation sites are going to be different between me and grandma), but it'll be a primary focus for health-conscious testers.

It's worth bearing this in mind, should the design of those tests be "gamed", or the advertisement to favour frequent testing, for commercial purposes (the same way older lightbulbs weren't 100% Argon - the main inert gas used IIRC - to prevent them lasting forever and ensuring a reliant customer base formed).


It'd be interesting to do a study testing epigenomes before and after histone deacetylase inhibitors.

MacUalraig
05-30-2019, 11:59 AM
Consumer testing of the epigenome will become a very lucrative market moving forwards.

Unlike your genome, the epigenome (the series of chemical signals that modify gene activity either directly or indirectly) is subject to change within the individual's lifetime. If I went from eating low-grade McDonald's cheeseburgers daily for a decade and then began daily meditation, exercise and a high vegetable diet, that'll reflect in the epigenome.

I can't see how epigenome testing would be useful for ancestry purposes (too many of those methylation-demethylation sites are going to be different between me and grandma), but it'll be a primary focus for health-conscious testers.

It's worth bearing this in mind, should the design of those tests be "gamed", or the advertisement to favour frequent testing, for commercial purposes (the same way older lightbulbs weren't 100% Argon - the main inert gas used IIRC - to prevent them lasting forever and ensuring a reliant customer base formed).

There seems to be a lot of research into early detection of diseases eg some cancers via methylation testing which could prove very beneficial.

The business model at Chronomics is based around a subscription which renews annually, you can't just buy one test. You can of course cancel before the first anniversary. Most of the controversy seems to centre around whether the information you get is 'actionable' but I might let it run just to see what changes there are.

There is a very detailed questionnaire which covers issues like how close to busy roads you have lived in case pollution has impacted you. I'm not sure how usable the information is though without incredibly detailed historical data. For example I grew up right up next to a dual carriage way but apart from the issue of how much data they have on that road/roads in general in the 60s and 70s there was a high hedge which I suspect absorbed the bulk of any pollutants even though we were playing outside near it a lot of the time.

Also I didn't especially want to give them my entire life's worth of postal addresses so I haven't filled it in. :-)

MacUalraig
06-22-2019, 01:47 PM
Apparently my epigenome results are ready, I've booked an online session for Wednesday to go through them. Ooh the ancticipation! :)

(looks like the WGS bit is still not back from the lab though).

MacUalraig
06-26-2019, 10:56 AM
I've had my initial results presentation with Chronomics. Good news - I'm only 50!! Young for my ago by 6 years :biggrin1:

So my 'biological age' is 50 but there is also detailed 'chromosome aging' based on the whole genome not just the CpG sites they use for the main figure. On this page 227/263 regions were 'younger than 52'.

"Chromosome Ageing
Biological ageing is a phenomenon that isn't restricted to the CpG positions used in the Biological Age predictor, it is something that happens throughout your genome."

and there's a nice little graphic to show these ages on a per chromosome basis.

It also reports 'Good' status for smoke exposure and metabolic status.

Not sure what to make of this level of info if anything. I'm more interested in the site specific methylation data.

In terms of what the testing actually consisted of, I'd assumed it was a methylation array test but in fact it was enrichment bisulfite sequencing with NovaSeq 100bp paired end reads. 5.3M sites high coverage then a further 20M sites at lower coverage. So there is a bisulfite BAM and some reports which they are going to upload. The separate WGS I ordered is due in 'at the end of the week' so they are going to put the whole lot up somewhere for me to download and play with. He started steering me towards a gVCF but I asked for either fastq and/or hg38 aligned BAM so should be getting that - they are on hg38.

The sex chromosomes were excluded from the analysis but the methylation raw data is available.

31184

31185

warwick
06-29-2019, 08:35 PM
I've had my initial results presentation with Chronomics. Good news - I'm only 50!! Young for my ago by 6 years :biggrin1:

So my 'biological age' is 50 but there is also detailed 'chromosome aging' based on the whole genome not just the CpG sites they use for the main figure. On this page 227/263 regions were 'younger than 52'.

"Chromosome Ageing
Biological ageing is a phenomenon that isn't restricted to the CpG positions used in the Biological Age predictor, it is something that happens throughout your genome."

and there's a nice little graphic to show these ages on a per chromosome basis.

It also reports 'Good' status for smoke exposure and metabolic status.

Not sure what to make of this level of info if anything. I'm more interested in the site specific methylation data.

In terms of what the testing actually consisted of, I'd assumed it was a methylation array test but in fact it was enrichment bisulfite sequencing with NovaSeq 100bp paired end reads. 5.3M sites high coverage then a further 20M sites at lower coverage. So there is a bisulfite BAM and some reports which they are going to upload. The separate WGS I ordered is due in 'at the end of the week' so they are going to put the whole lot up somewhere for me to download and play with. He started steering me towards a gVCF but I asked for either fastq and/or hg38 aligned BAM so should be getting that - they are on hg38.

The sex chromosomes were excluded from the analysis but the methylation raw data is available.

31184

31185

I really don't think this science is ready for prime time. I really doubt they have the evidence to back up their claims. Epigenetics research is very incomplete.

(FGC - affiliated)

warwick
06-29-2019, 09:21 PM
If they submitted this to a clinical service for validity they would get an "F."

Amerijoe
06-29-2019, 09:53 PM
Asked the following question. Hi: How does your testing of biological age compare to a telomere length report and itís correlation to biological aging.
Joe.

Hi: How does your testing of biological age compare to a telomere length report and itís correlation to biological aging.

Joe.
On Wed, Jun 26, 2019 at 12:16 PM, "Toby Call" <[email protected]> wrote:
Great question, the Epigenetic aging clock is far more accurate and biologically relevant than telomeres. Itís been called the new Ďgold standardí for measuring the aging process

Are you a scientist or health professional by any chance?

The amazing thing about Epigenetics is it gives far more insight into our health as a whole, rather than one dimensional telomeres.

Answer a little light in detail.

warwick
06-29-2019, 09:57 PM
Asked the following question. Hi: How does your testing of biological age compare to a telomere length report and it’s correlation to biological aging.
Joe.

Hi: How does your testing of biological age compare to a telomere length report and it’s correlation to biological aging.

Joe.
On Wed, Jun 26, 2019 at 12:16 PM, "Toby Call" <[email protected]> wrote:
Great question, the Epigenetic aging clock is far more accurate and biologically relevant than telomeres. It’s been called the new ‘gold standard’ for measuring the aging process

Are you a scientist or health professional by any chance?

The amazing thing about Epigenetics is it gives far more insight into our health as a whole, rather than one dimensional telomeres.

Answer a little light in detail.


little light in detail = nonsense

Incidentally, there is no clinically valid measure of your actual age. Anyone who claims they can give you a "true assessment" of your medical age is selling you a line of ....

FGC-affiliated

Amerijoe
06-30-2019, 03:14 AM
= nonsense

Incidentally, there is no clinically valid measure of your actual age. Anyone who claims they can give you a "true assessment" of your medical age is selling you a line of ....

FGC-affiliated

Since I lack in-depth knowledge in this particular area, would you please expound and provide the research if any which has led you to this opinion. Thanks in advance.

warwick
06-30-2019, 03:27 AM
Since I lack in-depth knowledge in this particular area, would you please expound and provide the research if any which has led you to this opinion. Thanks in advance.

There's no objective measurement of your medical age. If you go to a doctor, then the doctor may tell the patient, your signs are good for a person of your age but the notion that this can be computed from an epigenetic measurement in isolation from all types of other measurements, such as blood sugar, cholesterol, heart condition, blood pressure, weight, etc. is fanciful at this time.

Someone could have a "young" or "old" epigenetic status and have unhealthly stats for any of those other measurements (just as an example). It's misleading. Telomere measurements are also abysmal as a signal of aging. Based on recent research, telomere length accounts for less than 5% of the variability associated with aging. It just isn't a useful measurement at this point.

These are all "entertainment" tests at this point. It is quite likely that in ten or fifteen years epigenetics may prove more useful, but to infer a specific number for your age from an epigenetic test in isolation from all types of standard medical tests is malpractice and misleading.

example:
Source for the excerpt below:
https://blogs.plos.org/dnascience/2018/07/12/telomere-testing-science-or-snake-oil/


That’s why Mary Armanios, MD, from the Telomere Center in the McKusick-Nathans Institute of Genetic Medicine at Johns Hopkins University School of Medicine wrote “Telomeres in the Clinic, Not on TV” to accompany the case report and review article in the Mayo Clinic Proceedings. “These products present an oversimplified view of telomere length health: ‘short telomeres are bad’ equitable with aging, while ‘long telomeres are good’ and signify youthfulness,” she warns. Youthfulness of cancer cells, perhaps.

Limitations of measuring telomere lengths are many. The techniques used vary greatly and have poor reproducibility. Plus, DTC tests tend to report only the median or mean value for a patient’s submitted cells, which may only be meaningful for the lower extreme, and doesn’t include the distribution. Dr. Armanios compares this averaging shortcut to the “absurdity of reporting any value that is below the median for a white blood cell count as abnormal.”

In short, telomere lengths are too variable within a population, too variable within an individual, and too sensitive to environmental factors, to offer any reliable information for common disease risk. It also doesn’t appear to be individualized enough for another potential application, forensics.


Source for the below:
https://www.whatisepigenetics.com/epigenetics-avoiding-the-pull-of-pseudoscientific-nonsense/

In the same vein, it is imperative to present information reliably and accurately to the public without misleading those who may not be as up-to-date on the scientific details of a field like epigenetics. Social media and similar online outlets are notorious for spreading misinformation. Many rely on “clickbait” to entice readers as opposed to scientific substance, as the promise of exposure is too tempting to avoid for those who wish to spread hype and have no issue disregarding science. Fortunately, the scientific community seems to be pushing back against the assumptions many reporters, “experts”, and others are making about epigenetics.

warwick
06-30-2019, 04:02 AM
It is true that if an epigenetic test shows a very signficant difference between your chronological age and your "epigenetic" age, that could point to a disease or accelerated aging process. However, in the absence of other medical evidence, it is an incomplete measurement.

warwick
06-30-2019, 04:05 AM
It is true that if an epigenetic test shows a very signficant difference between your chronological age and your "epigenetic" age, that could point to a disease or accelerated aging process. However, in the absence of other medical evidence, it is an incomplete measurement.

However, a young or old "epigenetic age" is misleading without the context of other measurements. It can provide either false confidence that one is in good shape or false alarm that one is at risk.

Amerijoe
06-30-2019, 06:30 PM
It is true that if an epigenetic test shows a very signficant difference between your chronological age and your "epigenetic" age, that could point to a disease or accelerated aging process. However, in the absence of other medical evidence, it is an incomplete measurement.

I concur, at this point you are more right than wrong. Huge amounts of money are pouring into the science of aging and will only increase as the population ages. My point of interest are individuals suffering from PTSD due to the preadolescent stress conditions resulting in measurable epigenetic effects, one being increased telomere shortening.

Science does march on and someday, who knows, a drop of blood may unfold your life story right up to your final lights out.

warwick
06-30-2019, 10:47 PM
One gene that has proved very important in aging:

"Here, using a panel of 18 rodent species with diverse lifespans, we show that more robust DNA double-strand break (DSB) repair, but not nucleotide excision repair (NER), coevolves with longevity. Evolution of NER, unlike DSB, is shaped primarily by sunlight exposure. We further show that the capacity of the SIRT6 protein to promote DSB repair accounts for a major part of the variation in DSB repair efficacy between short- and long-lived species. We dissected the molecular differences between a weak (mouse) and a strong (beaver) SIRT6 protein and identified five amino acid residues that are fully responsible for their differential activities. Our findings demonstrate that DSB repair and SIRT6 have been optimized during the evolution of longevity, which provides new targets for anti-aging interventions."

https://www.cell.com/cell/fulltext/S0092-8674(19)30344-7?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com %2Fretrieve%2Fpii%2FS0092867419303447%3Fshowall%3D true

MacUalraig
07-01-2019, 08:11 AM
My main interest is getting the site level methylation reports/BAM which I'm still waiting for. I treat the headline result with some caution but its still interesting. The guy who runs the firm is Dr. Tom Stubbs who did this paper (on the mouse)

https://genomebiology.biomedcentral.com/articles/10.1186/s13059-017-1203-5

The paper has had 65 citations so far

http://citations.springer.com/item?doi=10.1186/s13059-017-1203-5

warwick
07-01-2019, 04:39 PM
My main interest is getting the site level methylation reports/BAM which I'm still waiting for. I treat the headline result with some caution but its still interesting. The guy who runs the firm is Dr. Tom Stubbs who did this paper (on the mouse)

https://genomebiology.biomedcentral.com/articles/10.1186/s13059-017-1203-5

The paper has had 65 citations so far

http://citations.springer.com/item?doi=10.1186/s13059-017-1203-5

Looks like a lot more mice will learn about their age from that paper.

MacUalraig
07-05-2019, 12:55 PM
Now managed to coax at least some of my raw data out of Chronomics, they used Bismark which comes from the same group who wrote FastQC.

https://www.bioinformatics.babraham.ac.uk/projects/bismark/

The file produced is the coverage file from the Bismark methylation extractor step:

"bismark2bedGraph also writes out a coverage file (using 1-based genomic genomic coordinates) that features two additional columns:
<chromosome> <start position> <end position> <methylation percentage> <count methylated> <count unmethylated>"

https://rawgit.com/FelixKrueger/Bismark/master/Docs/Bismark_User_Guide.html#iv-bismark-methylation-extractor

so in one example I have

chr1 redacted-pos redacted-pos 94.737 18 1

Not looked in great detail but I seem to have a lot which are heavily methylated but not 100% typically 75-9x%.

Total chr* rows is 19758384 but there are some separated out for the HLA at the top (1352 rows). Former number is roughly the total number of CpG sites in the genome.

warwick
07-28-2019, 07:02 PM
Now managed to coax at least some of my raw data out of Chronomics, they used Bismark which comes from the same group who wrote FastQC.

https://www.bioinformatics.babraham.ac.uk/projects/bismark/

The file produced is the coverage file from the Bismark methylation extractor step:

"bismark2bedGraph also writes out a coverage file (using 1-based genomic genomic coordinates) that features two additional columns:
<chromosome> <start position> <end position> <methylation percentage> <count methylated> <count unmethylated>"

https://rawgit.com/FelixKrueger/Bismark/master/Docs/Bismark_User_Guide.html#iv-bismark-methylation-extractor

so in one example I have

chr1 redacted-pos redacted-pos 94.737 18 1

Not looked in great detail but I seem to have a lot which are heavily methylated but not 100% typically 75-9x%.

Total chr* rows is 19758384 but there are some separated out for the HLA at the top (1352 rows). Former number is roughly the total number of CpG sites in the genome.

I think the main problem with this test is that it is very misleading. If a customer infers that their real age is "X" that is not correct. The customer may infer their life expectancy from the test. This test can't predict risk from other factors or diagnose disease, such as cancer, heart disease, and so forth.

Alternatively, if a customer gets a result that states they are much older than they expect, they may be discouraged or undergo mental stress.

It's just not responsible to offer this.

MacUalraig
08-12-2019, 02:10 PM
I think the main problem with this test is that it is very misleading. If a customer infers that their real age is "X" that is not correct. The customer may infer their life expectancy from the test. This test can't predict risk from other factors or diagnose disease, such as cancer, heart disease, and so forth.

Alternatively, if a customer gets a result that states they are much older than they expect, they may be discouraged or undergo mental stress.

It's just not responsible to offer this.

Your post doesn't relate to mine at all so I fail to see the point of quoting it. Illumina offer epigenetic testing, are they irresponsible too? Do you think my raw data is incorrect? I don't have any reason to doubt it.

There is no mention of biological or any other kind of age in my quoted post.