30 replies to this topic

[alito]

There was a thread back in 2014 regarding whole genome sequencing (http://www.longecity...nome-sequenced/). I expect whole genome sequencing to still be very expensive, but I'd think exome sequencing to have come down to mid three-figure prices.

 

I followed some of the links from the other thread to see if I could find prices. It looks like Gene by gene is nicely setup for consumers, accepting samples by mouth swab. They charge $1,095 (US$ I assume), which is a bit more than I expected but nothing outlandish.

 

BGI had an "Autumn promotion" doing exome sequencing for $250, which sounded great, but they haven't responded to my requests to see if they would accept spit instead of blood. Getting blood taken without it being requested by a doctor is hard in Australia and I don't think I'm game enough to do it myself. It also wouldn't be trivial to get the blood to Hong Kong in a usable form. They don't particularly seem setup to take consumer samples.

 

I haven't written to Macrogen (the Korean company mentioned in the other thread) or CD genomics to ask for prices yet.

 

Has anyone had any success getting any of these or any others to perform exome sequencing for you? Mag1 mentions in the other thread that they did but it wasn't any of those and it seemed to involve meeting with doctors. Any that are doable "over the net"?

 

Thank you

 

 

[niner]

This raises the question of what exome sequencing would get you that you couldn't get from a combination of a 23andMe-style broad SNP scan combined with imputation.  (I think there are a couple free imputation servers on the net-- mag1 has posted about it.)

[alito]

It'd get you rare and de-novo mutations. While there are much less of these, they are more clinically important, since otherwise they'd become common mutations.

 

http://massgenomics....-mutations.html says 38 de-novo mutations and 1.2% exonic, so on average 0.4 de novo mutations. It also claims 2/3 of genetic diagnoses are due to de-novo mutations.

 

That de-novo mutation rate also means that rare mutations should actually be quite common (ie a long tail), since the 0.4 per generation would accumulate from both sides.

 

The disadvantage is that you have almost no information on what those mutations do, but since they are on the exons you can tell what amino-acid they change and make guesses as to how important that would be.

[niner]

The disadvantage is that you have almost no information on what those mutations do, but since they are on the exons you can tell what amino-acid they change and make guesses as to how important that would be.

 

Maybe in some favorable cases, but a lot of things have to line up if you want to do that.  You need to know what the protein does, and you need to know the structure of it or of a near homolog.  Even with that, you'd only get an idea about whether the mutation was in the active site or not; there's a lot left to chance here.  Sometimes a mutation that you'd think would really hurt is rescued by a compenatory change in structure that you didn't expect, or a mutation that you'd think would be benign causes a critical change in the stability of the unfolded state... It's a hard problem.

[alito]

Yes, I basically agree. It'd give you a lead though.

 

The other thing an exome provides is information, which for some of us tends to be a intrinsic good (especially when its about oneself), regardless of its inapplicability. Like following the news.

[ymc]

Macrogen can do whole genome at around US$1500 but the problem is do you know how to interpret the result?

[alito]

Thanks for the data point. Do you happen to have the details? (Coverage and reads length? Is that $1500 straight from tissue to FASTQ? Do they take saliva or mouth swabs?)

 

I work in the bioinformatics field. While I can't interpret it in the "I know what this mutation does" sense, I know how to play around with it and know where to look for more information.

[ymc]

Thanks for the data point. Do you happen to have the details? (Coverage and reads length? Is that $1500 straight from tissue to FASTQ? Do they take saliva or mouth swabs?)

 

I work in the bioinformatics field. While I can't interpret it in the "I know what this mutation does" sense, I know how to play around with it and know where to look for more information.

 

They have HiSeq X, so it will be 2x150. The catch is you have to have at least 8 samples to fill out 8 lanes of one flowcell. This set you back $12,000 at least. 

 

They can process saliva for sure but need to add $50 processing fee.

 

I think you should email them for the latest quote. I suppose it is cheaper now.

 

Good luck!

[alito]

I'll email them. Thanks for the info. Getting 8 people may be tricky but maybe there's enough interest in this forum to make it happen.

[albedo]

Genos's services might be interesting to follow: "...Consumer genomics startup, Genos, is expanding its beta program and inviting Bio-IT World readers to sequence their whole exome at CLIA-certified 75x coverage for $399..."

http://www.bio-itwor...enos-model.aspx

[mag1]

I am so glad that this thread has picked up where the other one left off.

Clearly we are now moving closer and closer to the true unlocking of the Human Genome.

Apparently this will only happen when large numbers of the public have their exome sequenced and uploaded to genetic sites such as GEDmatch.

 

Up until now the human genome has remained largely undecipherable.

Most of the corporate and government databases have been closed to the public.

 

What I think is now on the horizon is a phenotype GEDmatch service for exome and genome scans.

This would be simply awesome!

The flood gates should open when we hit the $100-$200 exome price point.

 

I have spent years trying to figure out what all those 60,000 mutations mean.

It has been almost impossible.

 

With a phenotypic GEDmatch people could immediately uncover the genetics behind their traits.

This would be such a watershed moment for humanity.

As it is now the genetics research community is completely overwhelmed by the variation in human DNA.

If you go to dbsnp you will notice that nearly all the snps report "clinical significance unknown". 

When exome sequencing is as common as gene chipping is today, we should expect a rapid revelation of the meaning of DNA.  

 

 

Consider what might happen if I could upload my exome vcf file to GEDmatch II.

People who upload could disclose phenotypes: even if they simply disclosed that they did not have a particular

phenotype this would be valuable information. All the shared variants of normal phenotype could be removed

from the search space. This would be an extremely powerful tool. It would not be completely unexpected that one could upload

an exome file and simply push a button and instantly find which variants might be causing a particular trait of interest.

 

As it is now using only SNP gene chips instead of an exome file you never really know what variants you share with your

DNA relatives. Without having the shared variants you are never entirely sure whether a DNA relative who shares

an IBD segment with you also shares variants on this segment which might have occurred more recently than the

most common recent ancestor or which strand of DNA they are on. Furthermore with exome sequencing anyone who shared a

variant with you could add information no matter how small the shared segment. Most platforms discard matches that fall below

certain number of cMs. A high quality exome variant would not need to be discarded no matter how small the size of the shared

segment in cMs.     

 

Typically current genetic databases now have close to saturation coverage of the genome.

At the right price, exomes should finally unlock the human genome.

If none of the big players are interested in seeing this happen then it is pretty much up to We the People to move this one over the line.     

All we need to make this a reality is a pretty simple piece of software that searches for shared variants, a form that asks about phenotypes 

and people willing to step up and have their exomes sequenced.

This is one of those times where we need the massive to join together and create a network effect.

​Once we hit similar numbers for exomes as are now on gene chip sites (about 2-3 million), the entire human genome could be deciphered.  

  

Edited by mag1, 06 November 2016 - 05:29 AM.

[mag1]

Simply amazing.

​These nanopores are moving towards commercial viability.

 

On one of the other threads it was suggested that it would cost hundreds of thousands of dollars to sequence a human

genome with nanopores.

 

The url below is blueskying less than $1000!  

Yep, that could certainly make things interesting.

I was thinking the exome revolution might be a few years away. Edit that. A few months away?

 

I am not sure, though, nanopore technology might be just what we are looking for to move down the price point of exome sequencing.

Possibly these nanopores might allow you to do a 1x read until you hit an exon and then you could automatically move up to 75x coverage.

It is not totally clear to me what the true comparisons would be between typical Illumina sequencing and Nanopore sequencing in terms

of accuracy etc.

 

Of course being able to sequence your own genome with your own little Minion has to be worth something on the cool factor scale. I mean sure I would be willing to pay them a premium to Do it Myself (DIM). As long as they did not insist on truthful answers to all their intrusive questions about what University Research Lab I work at etc. . If we could move nanopore exome sequencing down to a $100 price that would be highly disruptive.

 

https://nanoporetech...n-genome-minion

Edited by mag1, 06 November 2016 - 07:01 AM.

[albedo]

Genos's services might be interesting to follow: "...Consumer genomics startup, Genos, is expanding its beta program and inviting Bio-IT World readers to sequence their whole exome at CLIA-certified 75x coverage for $399..."

http://www.bio-itwor...enos-model.aspx

 

In case your missed it, here is the recent announcement from Promethease:

 

"...We're pleased to announce that a new company, Genos, is now producing a reasonably-priced sequence-based genome service that tests the majority of the DNA variants in all of your genes, and they have worked with us to ensure Promethease compatibility.
Their current price for sequencing your whole exome is $399, but until Dec. 12, as a Promethease user you can purchase as many as you like for $359 (a $40 discount), and for each exome we will run a new Promethease report at no charge. To get this discount, you must use URL:

https://www.genosres...ail-GI&pid=prom
Discount will be reflected on the final Genos order page.

While we benefit from having more exomes to analyze, which helps us curate more variants for SNPedia, you benefit from having more comprehensive data. Additional advantages to having exome sequence are detailed here...." (bold mine)

 

[mag1]

albedo, I am getting excited!

 

Exome pricing is getting closer to the point where this will go mainstream.

Once this occurs it will be greatly easier to narrow down the variant files to mutants that have actual relevance.

 

Most of the big genomics players do not appear to want to play a constructive role.

Time for The People to step up and crack the genome.

 

I have been trying to understand a family member's exome for years.

With a few simple phenotype questions with the exome sequencing, it should be possible to have answers immediately.

 

   

Edited by mag1, 25 November 2016 - 12:50 AM.

[albedo]

Yes, mag1, its getting closer and affordable for everyone. This is part of the personalized medicine momentum, people are asking and MDs would need to be trained for interpretation. I was at a very interesting talk by Prof. Michael Snyder (Stanford) at University of Geneva, he was one of the fist to apply a very in depth iPOP (integrated Personal Omics Profiling) to his own data. Maybe you can check out also these links (incl. Veritas Genetics offering for full genome):

 

https://mediaserver....e.ch/play/97787

https://www.ncbi.nlm...pubmed/22424236

https://www.veritasg...cs.com/mygenome

 

[mag1]

albedo, do you have any idea whether Genos would provide the full exome data set with the reads?

Seeing the data can help spot errors that the software does not always intrepret correctly.

 

Also like to know whether Genos might offer phasing with their service. There have been

products introduced that perform this with long reads and bar coding. 

 

This looks like a very good deal from Genos. I would love to sign up if I could arrange for a

few of these extras.

[mag1]

I do not think it is fully appreciated how totally governments have betrayed their people regarding genomics.

Governments have conducted a large amount of population based sequencing.

Virtually none of this dataset has made it out to open access platforms.

 

When are people going to push back against this?

This is just another example of genetic neocolonialism.

All these genetic samples were freely given by people, likely with the intention of allowing fair access.

For example, if someone gave a DNA sample and they included the information that they did not have illness X in their family, then

there should be no reason why this information could not be disclosed in a publicly accessible database of exome variants.

 

Perhaps the ability to disclose genetic results by research participants should be considered a human right.

A human right that is currently being denied these people.

 

These studies were, of course, funded by We the People.

It is We the People who spent our money.

Why do they take our money, do research, and then not provide the results?

No show, no dough.

 

It would already be nearly automatic to interpret an exome or genome.

When will government be held to account?

 

Open science should be the default, not the exception.

   

Edited by mag1, 27 November 2016 - 09:37 PM.

[alito]

albedo, do you have any idea whether Genos would provide the full exome data set with the reads?

Seeing the data can help spot errors that the software does not always intrepret correctly.

 

I asked about FASTQ availability and they said to ask for them once you get the results and they would provide them.

 

 

Also like to know whether Genos might offer phasing with their service. There have been

products introduced that perform this with long reads and bar coding. 

 

Seems highly unlikely given the price point. They are most likely going with whatever provides the lowest price per read at the moment. (I assume you are referring to 10X's technology, which isn't cheap)

[mag1]

alito, thank you for your response.

 

I was somewhat confused when you mentioned the FASTQ file previously.

The FASTQC file for our first exome sequence was small; only about a meg.

The really massive file of over 5 gigs was the bam file.

 

This was the file that I transferred from our Illumina base space account to genomic applications.   

I think this is the one with the reads.

I would want the read file from Genos.

 

It is true that it would be much easier for them simply give you the VCF file.

However, I would want to know the coverage and I would also want to see the reads because often you can

see things that the software gets wrong. I don't see why Genos would have a big problem with sharing this information

to their clients. They are in a data intensive industry so they might as well be ready. Data storage is no longer greater expensive.

Perhaps they could simply push it to the cloud. Some of the genomics providers do not have much of a problem hosting these

huge datasets for considerable lengths of time.

 

​Yes, you are quite right about the expense of the 10x's technology.

However, I was thinking more about IGenomX's new technology.

From what I understand they are moving to near zero incremental cost.

I am also interested in the possibility of having a more uniform coverage of the exome.

The IGenomX's product also appears to help with that deficiency in typical sequencing.   

 

It was somewhat disappointing in our first exome sequence to see some areas read sometimes 500 or more times and other areas not read at all.

500 reads at a locus is completely empty sequencing. There is zero marginal value at even much over 100 times coverage. Any technology that

could more evenly distribute the reads over the entire exons would be such a bonus.   

 

 

 

Edited by mag1, 28 November 2016 - 12:16 AM.

[mag1]

The IGenomX does of course require a quarter million dollar Thunderstorm, though considering its high throughout this would not be

an overly large burden for a large scale sequencer. 

Edited by mag1, 28 November 2016 - 12:15 AM.

[alito]

alito, thank you for your response.

 

I was somewhat confused when you mentioned the FASTQ file previously.

The FASTQC file for our first exome sequence was small; only about a meg.

The really massive file of over 5 gigs was the bam file.

 

This was the file that I transferred from our Illumina base space to genomic applications.   

I think this is the one with the reads.

I would want the read file from Genos.

I think you are confusing FASTQ and FASTQC files. FASTQ files are the files with the reads (and quality scores). FASTQC is just a QC (quality control) file summarising the reads. BAM files are alignment files listing the alignment of the reads from FASTQ files to a specific genome. You can (mostly) get the FASTQ files back from the BAM files, but I'd rather have the FASTQ files itself since the alignment depends on what aligner you are using and what version of the genome.

 

I don't see why Genos would have a big problem with sharing this information to their clients.

 

Like I mentioned, I don't think they do have a problem. I think it's more of a matter of most people not knowing what to do with raw files, so they might as well just provide something smaller and easier to understand.

​

Yes, you are quite right about the expense of the 10x's technology.

However, I was thinking more about IGenomX's new technology.

 

Ah I see. I don't know anything about IGenomX. I'll wait till I see data.
 

 

It was somewhat disappointing in our first exome sequence to see some areas read sometimes 500 or more times and other areas not read at all.

500 reads at a locus is completely empty sequencing. There is zero marginal value at even much over 100 times coverage

 

 

I think you just had a bad experience. Most exomes I see have pretty good coverage of the target regions (>99%)

[mag1]

That's great! I wanted to make sure that it was the read file that Genos would report to you.

I have checked in my basespace account and I did not see the FAST files only the FASTQC.

​I suppose the FAST files must be there somewhere.

The only big ones I found was BAM.

 

I took the large files (FAST?) and uploaded to Genome Browser and it was great.

You could see all the reads and you could also see where the aligner might have had some trouble

in a highly repetitive section. The barcoding services such as 10X would at least make sure that

you aligned the sequencer in the right genomic neighbourhood. As it is, sometimes you have the impression

that they simply have to force the read where it really doesn't belong, sort of like in a jig saw puzzle when you

just have to bang in a piece with a hammer if necessary.     

 

It is really great that exome sequencing is moving to a commodity product.

Buy it and don't worry.

At some point ideally, you could order an exome sequence that was phased with long reads at 100X uniform coverage

and perhaps a 1000 base pair buffer from the ends of each gene as there can be interesting segments just beyond

the exons themselves. 

 

Things were still being worked out a few years ago when we had it done.

 

Could anyone confirm that the exomes from Genos have good uniform coverage?

What is the minimum coverage in an exon?  

Also like to know whether mitochondria DNA is included.

 

In our first scan, I had not realized that the mitochondria had been sequenced until I tried out the IGV program on basespace.

Surprisingly there it was. The other programs including Genome Browser did not report on the mtDNA.

mtDNA is of considerable importance for health, so it is worthwhile to know whether it has been sequenced or not.

In fact, there are usually often quite a few mitochondria in cells, so the coverage for mtDNA can be very high.  

 

[albedo]

albedo, do you have any idea whether Genos would provide the full exome data set with the reads?

Seeing the data can help spot errors that the software does not always intrepret correctly.

 

Also like to know whether Genos might offer phasing with their service. There have been

products introduced that perform this with long reads and bar coding. 

...

 

My apologizes mag1, just seen. I do not have the reply but see alito gave one in the meantime. Very interesting.
 

[mag1]

This Genos price seems too good to miss.

I am sitting on fence, though it would not take much to get me onboard.

 

75x genome with uniform coverage and the read file sounds very good at $350.

The phasing would not even matter because we would have exome sequences for both parent and offspring.

This should give us a phased result at no expense.

Also the mitochondrial DNA for the offspring should be the same as for the parent, so we would have mtDNA as well.

 

Obviously the price wants to move down and it probably will be $100 or $200 cheaper a year from now, but why wait?

 

Can anyone confirm that Genos' exome scans do in fact have uniform coverage across exons?

Edited by mag1, 29 November 2016 - 02:55 AM.

[mag1]

I couldn't miss out on the deal that Genos was offering!

Did due diligence, everything checked, signed up, and the kit should be here any time now.

 

This is a very good price.

Technical support noted that with parent offspring exome sequencing one would get phasing as a bonus at no expense.

 

This is like an early Christmas present!

Very Excited! 

 

[mag1]

Something that I find very surprising is that on the order ticket Genos gives everyone an order number.

The order number appears to be the total numbers they have processed since they started up.

I am around 1000.

 

What I find troubling is: Have only 1000 people stepped up to order their exome?

Genos is one of the first consumer friendly exome sequencing services out there.

When we did an exome sequence a few years the company had to really think about risking taking our business.

Genos now has it all in the open and people have still not lined up.

 

This is difficult to understand.

There are more than 2 million births every year in the US.

Every one who is contemplating becoming a parent simply must get exomed.

 

Exoming both parents would give all the recessive homozygous that there child might inherit.

Why would parents want to just wait and see what the genetic roulette wheel came up with?

Recessive illnesses can have severe health consequences and would likely never have been

present in a family before. It is very difficult to understand why people would not now be flocking

to get exomed.

 

 

 

 

 

 

 

  

 

 

 

 

  

[mag1]

Could anyone help out?

Would it be possible to send a lab a DNA sample and let them work through whatever might be genetically defective with the cells?

 

 

I was thinking that they could take the cells and then grow organoids or perhaps have some way to grow the cells to see whether they

exhibit a health condition. The idea would be to remove half of the chromosomes in a cell and replace these chromosomes with healthy chromosomes. This cell could be grown into an organoid etc. If this organoid were to have a disease phenotype, then it would be known that the half of chromosomes that had remained in the cell contained the disease variant of interest.

 

Half of the remaining original chromosomes could then be removed and healthy chromosomes could be replaced again. In this way after

5 such replacements there would only be a single chromosome that could house the disease variant. Perhaps even a single chromosome could be snipped and a healthy chromosome added to it. It might be possible to only search a small part of the genome where a disease causing variant might be present.

 

Not sure why this approach does not appear to have been used to date. Most of the research community seems very focused on GWAS studies even though the typical hit with GWAS has only a slightly increased risk of a trait.  

 

Edited by mag1, 14 December 2016 - 01:42 AM.

[albedo]

..

This is difficult to understand.

There are more than 2 million births every year in the US.

Every one who is contemplating becoming a parent simply must get exomed.

..

 

Hi mag1! I liked your posts and your enthusiasm! Thanks. Just a comment I had reading this post is the quality of whole exome (WES) vs whole genome (WGS) sequencing. While the cost of latter is going down its quality is better than in the former, even most importantly when contemplating becoming a parent. There are technical reasons for this e.g. see here:

 

Belkadi A, Bolze A, Itan Y, et al. Whole-genome sequencing is more powerful than whole-exome sequencing for detecting exome variants. Proc Natl Acad Sci USA. 2015;112(17):5473-8.

http://www.pnas.org/...ent/112/17/5473
 

[mag1]

Albedo, great to have at least someone in my fanclub.

 

Sorry I have not got back to your post earlier.

 

I am sure everyone on this thread has some appreciation of how extremely important this rapidly approaching wave of

genetic change will be for humanity.

 

The technology is now already in place.

Once the price for consumer sequencing moves down even somewhat from current levels, the human genome should quickly unlock.

It is probably a very bad idea to make predictions though within 5 years gives quite a bit of padding on the backend.

 

This is close.

It will profoundly change humanity.

 

I wonder if medicine, as it is now defined, might even have a future that stretches much past the mid-term time horizon.

The discussion probably should start sooner than later.

In a genetically engineered future, illness would become greatly less common and perhaps nearly nonexistent at younger ages.

 

Yet, it saddens me to realize that our community will likely soon be divided into the adopters and the rejecters of genetic technology.

This will be a far more important division than wealth or other categories.

 

It seems all too clear that some people will want to be as they always have been.

They will want to reject genetic selection just as others have tried to deny evolution.

We could wind up with a society split based on the freedom to choose to live in the future or the past.

 

At the same time this has a personal dimension. My family has a range of inherited illnesses such

as autosomal dominant Alzheimer's, polyp/colorectal cancer syndrome, probably several others.

 

I could live without cancer, probably also cardiac arrhythmia and others though I am not sure about those traits that directly relate to the functioning of the brain..

 

I am not sure whether I would want not to have Alzheimer's in my life. Alzheimer's is our way of life. My family has spent decades caring for family members with dementing illness. It has been such a blessing to be a caregiver. It

has been the central focus of our life. How could you live without it? So many people without such a focus

just drift through life. Being a caregiver, forces you to learn to care. In our increasing sanitized modern world

this life skill appears to have become a muscle not exercised.   

 

Perhaps society will have to carefully consider such issues. Humanity might have been designed through

a natural selection process in which the optimal genetics for a community included those with illnesses.

I am becoming increasingly worried about the simple minded notion that we will all be so much happier

when we live eugenically.

 

Our entire society would be disrupted if the care receiver and care giver no longer existed. Close to the entire

structure of our life is related to these roles. 

 

 

 

   

Edited by mag1, 09 February 2017 - 01:09 AM.

[world33]

genos.co for 499$; recommended by a guy from snpedia.com