12 ms·

The era of fast, cheap genome sequencing is here

88 points by ColinWright 4y ago

mattnewport 4y ago
And yet the era of being able to do anything really useful with the information has yet to arrive.
- cjbgkagh 4y ago
  That future is just not evenly distributed yet.
  - jjtheblunt 4y ago
    that's an excellent way to put it (seriously)
- pfdietz 4y ago
  Sequencing cancer cell genes is quite common now, to determine just what kind of cancer is there and which treatments are appropriate.
  - dominotw 4y ago
    GP probably meant for completely healthy people who are the target audience for doing these . Also the cancer drugs like PARP inhibitors seem to be helping everyone regardless of mutation status in clinical trials. So the connection is much less clear.
    - pfdietz 4y ago
      There are many cancer drugs that target mutations in specific proteins. Just what treatment will be given for a cancer often depends on what's been mutated.
      - dominotw 4y ago
        > Just what treatment will be given for a cancer often depends on what's been mutated. What I was saying was that those drugs have been proving effective regardless of mutation status. So mutation x = give drug Y connection is not really well understood. Example, olaprib for Prostate cancer > the addition of olaparib was associated with a significantly prolonged radiographic PFS irrespective of HRR status https://www.practiceupdate.com/content/asco-gu-2022-propel-suggests-adding-olaparib-to-abiraterone-improves-outcomes-in-castration-resistant-prostate-cancer-regardless-of-homologous-recombination-repair-status/132359 https://www.practiceupdate.com/content/asco-gu-2022-propel-s...
        pfdietz 4y ago
        > What I was saying was that those drugs have been proving effective regardless of mutation status. I don't believe this is the case in general. Indeed, all cancer drugs fail sometimes, and there has to be something different about the cancers they fail on. If not the mutation status, then what?
        dominotw 4y ago
        I am not quite sure how to respond to 'I don't believe' response. Even FDA is approving these mutation targeting drugs without mutation status. Niraparib was approved for Ovarian Cancer even if you don't have HRD mutations. Approval was based on NOVA trial, you can lookup the details. I am not saying there isn't any connection, what I am saying is it not really well understood or clearcut as mutation x = drug y that general public believes about these drugs.
- Der_Einzige 4y ago
  Clustering, dimensionality reduction, frequent pattern matching can get you a lot of the way there.
- mbreese 4y ago
  Rare disease diagnosis, newborn screening, molecularly targeted cancer therapies… We’re starting to get some good utility from sequencing. Currently whole exome is more informative than whole genome. But, there are definitely some good clinical uses!
  - sacul 4y ago
    My daughter has a rare genetic disorder, cardiofaciocutaneous syndrome (CFC), and it was not until 2 years in that we were able to get a diagnosis after a whole gene sequence. It was game-changing to get the diagnosis, because it connected a to a community and to doctors who could help. https://rarediseases.info.nih.gov/diseases/9146/cardiofaciocutaneous-syndrome https://rarediseases.info.nih.gov/diseases/9146/cardiofacioc...
- FollowingTheDao 4y ago
  This is the biggest misconception. I have literally reverse my hyperlipidemia, hypertension, and been able to get off all of my meds for my schizoaffective Bipolar Disorder because of what I have learned from my genetics and changing my diet/lifestyle to suit what my body needs. Just because doctors do not know what to do with it (they are decades behind most research) it does not mean you cannot learn about this and take action on your own. Just knowing my FADS1, FADS2 genetics was enough to change my diet to seafood only and raise my HDL from a consistent 35 to over 55 in three weeks. Doctors put me on a statin which lowered my HDL to 25.
  - dominotw 4y ago
    > seafood only and raise my HDL Seafood helps everyone with HDL. Is 35 -> 55 not possible with those genes that you mentioned ?
    - FollowingTheDao 4y ago
      It was not only those genes, but omega 3 is crucial for what’s called reverse cholesterol transport. So I’m sure it might help anyone, I just need more help. https://pubmed.ncbi.nlm.nih.gov/28984832/ https://pubmed.ncbi.nlm.nih.gov/28984832/
giuliomagnifico 4y ago
This should be a good news, because with DNA test you can find some hereditary diseases, but I’m a bit skeptical about the privacy, in particular when I read things like this: “Mass DNA Collection in the Tibet Autonomous Region from 2016–2022” https://citizenlab.ca/2022/09/mass-dna-collection-in-the-tibet-autonomous-region/ https://citizenlab.ca/2022/09/mass-dna-collection-in-the-tib... The law will absolutely have to adapt to these changes.
- Dracophoenix 4y ago
  This will certainly put a damper on the sperm donation market.
  - quesera 4y ago
    An individual sperm cell does not contain a full genome. You cannot even reliably determine a donor's hair color, from a single sperm cell. I imagine it's much more difficult to reconstruct a full genome from a sperm donation than from, say, saliva on a discarded cup.
    - chromatin 4y ago
      This comment is misleading, and possibly wrong depending on what you mean by “full genome” (not an accepted technical term). A sperm cell contains one complete haploid genome.
      - quesera 4y ago
        You're right, it's not as simple as I described. But (my bio is rusty) a haploid genome (gamete) cannot fully represent the genome of a diploid human. If you're looking for genetic disorders, these will often be the result of recessive traits, right? So having only half of the chromosome sets is not indicative. That's my example of hair color. With a full genome sequence, you could determine the person's hair color, but with a haploid genome, you could not. If OP's joke is about the privacy implications of sperm donation, then saliva seems more risky.
nanomonkey 4y ago
My father died of pancreatic cancer this week after getting a diagnosis 36 days prior. Since then I've been on a minor crusade to find methods of making sure I have a better diagnosis of anything similar. Full body MRI can be had for ~$2k. I can imagine going in and getting a genome sequencing and an MRI scan done, and then over the years having a much better idea of the changes going on in your body as you age, along with an evolving map of what to look out for. It's exiting, and yet I'm worried about handing over information that could be used against me by corporations, the government or insurance companies.
- pavlov 4y ago
  Maybe get your MRI and genome done in Europe which has very strong protections for this kind of data?
  - bobf 4y ago
    Medical tourism and paying in cash wherever is convenient/inexpensive might also be an option to consider - there are a lot of places that will be cheaper than Europe.
- divbzero 4y ago
  I am very sorry for your loss. I know there are efforts like Galleri [1] to detect cancer early using liquid biopsy [2] but I don’t know how effective they are at the moment. [1]: https://www.galleri.com/ https://www.galleri.com/ [2]: https://en.wikipedia.org/wiki/Liquid_biopsy https://en.wikipedia.org/wiki/Liquid_biopsy
- zemvpferreira 4y ago
  My condolences for your loss. My own father is a physican and very against any elective exams himself. His reasoning is that his colleagues are excessively interventive so the less medicine, the better. I wish your own father had been diagnosed in time but I can see the downside in doing too many exams as well. Doctors won't be able to keep their hands off treating any little lump they find, even if it's to your detriment. Case in point, surgery for back pain causes said pains to worsen at about the same rate as it improves them.
derekja 4y ago
sequencing continue to drop in price far faster than moore's law. https://www.genome.gov/about-genomics/fact-sheets/DNA-Sequencing-Costs-Data https://www.genome.gov/about-genomics/fact-sheets/DNA-Sequen... Here is the archive of the wired article, btw: https://web.archive.org/web/20221001173347/https://www.wired.com/story/the-era-of-fast-cheap-genome-sequencing-is-here/ https://web.archive.org/web/20221001173347/https://www.wired...
- trention 4y ago
  Not if you look at the period after 2015 only.
- dekhn 4y ago
  sequencing stopped dropping in real prices 5+ years ago.
peter303 4y ago
The second identifiable person to have his genome sequenced- James Watson- had around 30 known genome defects at the time. However most of them did not manifest as disease. That is a mystery.
- FollowingTheDao 4y ago
  SNP changes in genes are not usually mutations that cause "defects". Most usually they increase or decrease risk for a disorder based on more of a polygenic profile.
- tejtm 4y ago
  For small enough values of defect, it is par for the course.
- dekhn 4y ago
  No, it's only a mystery for people who have a truly simplified version of genetics and its relationship to disease phenotypes.
sometimeshuman 4y ago
Can someone recommend where one can order a full genetic sequence currently in the US and get the raw data ? It seemed highly-regulated when I last checked and required it be order through your doctor. The article touches on this a bit how the price gets into the thousands due to markups and counseling. Also how many full sequences would I have to get before the data is statistically sound. Since there is likely a base pair or two that will be different between the cells (maybe) and the sequencer hardware and process will likely have errors too. My gut tells me 10 to 20 but perhaps that seems like overkill since many base-pairs might be junk/repeats/irrelevant anyway and the raw data will have some confidence indication as well IIRC.
- jghn 4y ago
  You should be looking for 30x coverage. There are a number of companies that will do 30x WGS and give you the BAM file. I believe Nebula is one of them, but know there are other names out there.
  - aroch 4y ago
    Nebula most likely ship your samples to China for processing and sequencing because it's so much cheaper. While that might not be a deal breaker for some, I would personally prefer my sample not be shipped elsewhere.
  - FollowingTheDao 4y ago
    Stay away from Nebula. I had such a hassle with them and after six months of complaining with mostly no response I had to threaten then with a civil suit and they finally replied "your sample was no good, here is your money back."
- chrisrohlf 4y ago
  I asked this same question not long ago and the CEO of a well known company in this same research area reached out and recommended Veritas Genomics https://veritasgenetics.com/mygenome/ https://veritasgenetics.com/mygenome/ I’ve yet to finish the order process but I share this here because in my research of companies providing full sequencing services I never came across them and yet they seem far more reputable than others. Current SEO for these services turns up a lot of sketchy companies.
gavinray 4y ago
It's not common knowledge, but there's an algorithm called "genomic imputation" It can take incomplete data, like the kind you get from 23andMe, etc, and magically "fill in the gaps" to compute the majority of the rest of your genome There used to be a free service https://impute.me https://impute.me that would do this for you https://twitter.com/lassefolkersen/status/1547637086349537284?t=ewGfTYY6DywjCwrMRGbWSQ&s=19 https://twitter.com/lassefolkersen/status/154763708634953728... Search Google Scholar for "Lasse Folkersen", he is active in the field and has a lot of great papers about this
bilsbie 4y ago
I recall this being predicted something like 10 years ago. I wonder why a slowdown happened.
haihaibye 4y ago
Cheap* *To produce the raw data, and does not include analysis It's likely the bill for a consultant (senior doctor) to examine a patient, order a test then read and sign off on a report is already more than the chemistry + cpu + analysis already.
pazimzadeh 4y ago
It's good news but it's not as great as many people used to think. Having a particular gene doesn't mean that it's actually turned on. Your heart and your hair cells have the exact same genome, but do vastly different things. I am not excited as I used to be about the promise of cheap genome sequencing by itself. It needs to be paired with lots of other data for it to have any use, except for determining phylogeny/ancestry. Even more exciting than this would be cheap DNA synthesis, allowing us to build and test complicated genetic circuits in fewer steps. The current methods usually rely on stitching together existing sequences, rather than creating what you need from scratch. The price of DNA sequencing has fallen much faster than the price of DNA synthesis. Here is a comparison of the cost of DNA sequencing vs synthesis: http://www.synthesis.cc/synthesis/2016/03/on_dna_and_transistors http://www.synthesis.cc/synthesis/2016/03/on_dna_and_transis.... It would be good to find more recent data though.

88 points by ColinWright 4y ago

mattnewport 4y ago

And yet the era of being able to do anything really useful with the information has yet to arrive.

cjbgkagh 4y ago
That future is just not evenly distributed yet.
- jjtheblunt 4y ago
  that's an excellent way to put it (seriously)
pfdietz 4y ago
Sequencing cancer cell genes is quite common now, to determine just what kind of cancer is there and which treatments are appropriate.
- dominotw 4y ago
  GP probably meant for completely healthy people who are the target audience for doing these . Also the cancer drugs like PARP inhibitors seem to be helping everyone regardless of mutation status in clinical trials. So the connection is much less clear.
  - pfdietz 4y ago
    There are many cancer drugs that target mutations in specific proteins. Just what treatment will be given for a cancer often depends on what's been mutated.
    - dominotw 4y ago
      > Just what treatment will be given for a cancer often depends on what's been mutated. What I was saying was that those drugs have been proving effective regardless of mutation status. So mutation x = give drug Y connection is not really well understood. Example, olaprib for Prostate cancer > the addition of olaparib was associated with a significantly prolonged radiographic PFS irrespective of HRR status https://www.practiceupdate.com/content/asco-gu-2022-propel-suggests-adding-olaparib-to-abiraterone-improves-outcomes-in-castration-resistant-prostate-cancer-regardless-of-homologous-recombination-repair-status/132359 https://www.practiceupdate.com/content/asco-gu-2022-propel-s...
      - pfdietz 4y ago
        > What I was saying was that those drugs have been proving effective regardless of mutation status. I don't believe this is the case in general. Indeed, all cancer drugs fail sometimes, and there has to be something different about the cancers they fail on. If not the mutation status, then what?
        dominotw 4y ago
        I am not quite sure how to respond to 'I don't believe' response. Even FDA is approving these mutation targeting drugs without mutation status. Niraparib was approved for Ovarian Cancer even if you don't have HRD mutations. Approval was based on NOVA trial, you can lookup the details. I am not saying there isn't any connection, what I am saying is it not really well understood or clearcut as mutation x = drug y that general public believes about these drugs.
Der_Einzige 4y ago
Clustering, dimensionality reduction, frequent pattern matching can get you a lot of the way there.
mbreese 4y ago
Rare disease diagnosis, newborn screening, molecularly targeted cancer therapies… We’re starting to get some good utility from sequencing. Currently whole exome is more informative than whole genome. But, there are definitely some good clinical uses!
- sacul 4y ago
  My daughter has a rare genetic disorder, cardiofaciocutaneous syndrome (CFC), and it was not until 2 years in that we were able to get a diagnosis after a whole gene sequence. It was game-changing to get the diagnosis, because it connected a to a community and to doctors who could help. https://rarediseases.info.nih.gov/diseases/9146/cardiofaciocutaneous-syndrome https://rarediseases.info.nih.gov/diseases/9146/cardiofacioc...
FollowingTheDao 4y ago
This is the biggest misconception. I have literally reverse my hyperlipidemia, hypertension, and been able to get off all of my meds for my schizoaffective Bipolar Disorder because of what I have learned from my genetics and changing my diet/lifestyle to suit what my body needs. Just because doctors do not know what to do with it (they are decades behind most research) it does not mean you cannot learn about this and take action on your own. Just knowing my FADS1, FADS2 genetics was enough to change my diet to seafood only and raise my HDL from a consistent 35 to over 55 in three weeks. Doctors put me on a statin which lowered my HDL to 25.
- dominotw 4y ago
  > seafood only and raise my HDL Seafood helps everyone with HDL. Is 35 -> 55 not possible with those genes that you mentioned ?
  - FollowingTheDao 4y ago
    It was not only those genes, but omega 3 is crucial for what’s called reverse cholesterol transport. So I’m sure it might help anyone, I just need more help. https://pubmed.ncbi.nlm.nih.gov/28984832/ https://pubmed.ncbi.nlm.nih.gov/28984832/

giuliomagnifico 4y ago

This should be a good news, because with DNA test you can find some hereditary diseases, but I’m a bit skeptical about the privacy, in particular when I read things like this: “Mass DNA Collection in the Tibet Autonomous Region from 2016–2022” https://citizenlab.ca/2022/09/mass-dna-collection-in-the-tibet-autonomous-region/ https://citizenlab.ca/2022/09/mass-dna-collection-in-the-tib... The law will absolutely have to adapt to these changes.

Dracophoenix 4y ago
This will certainly put a damper on the sperm donation market.
- quesera 4y ago
  An individual sperm cell does not contain a full genome. You cannot even reliably determine a donor's hair color, from a single sperm cell. I imagine it's much more difficult to reconstruct a full genome from a sperm donation than from, say, saliva on a discarded cup.
  - chromatin 4y ago
    This comment is misleading, and possibly wrong depending on what you mean by “full genome” (not an accepted technical term). A sperm cell contains one complete haploid genome.
    - quesera 4y ago
      You're right, it's not as simple as I described. But (my bio is rusty) a haploid genome (gamete) cannot fully represent the genome of a diploid human. If you're looking for genetic disorders, these will often be the result of recessive traits, right? So having only half of the chromosome sets is not indicative. That's my example of hair color. With a full genome sequence, you could determine the person's hair color, but with a haploid genome, you could not. If OP's joke is about the privacy implications of sperm donation, then saliva seems more risky.

nanomonkey 4y ago

My father died of pancreatic cancer this week after getting a diagnosis 36 days prior. Since then I've been on a minor crusade to find methods of making sure I have a better diagnosis of anything similar. Full body MRI can be had for ~$2k. I can imagine going in and getting a genome sequencing and an MRI scan done, and then over the years having a much better idea of the changes going on in your body as you age, along with an evolving map of what to look out for. It's exiting, and yet I'm worried about handing over information that could be used against me by corporations, the government or insurance companies.

pavlov 4y ago
Maybe get your MRI and genome done in Europe which has very strong protections for this kind of data?
- bobf 4y ago
  Medical tourism and paying in cash wherever is convenient/inexpensive might also be an option to consider - there are a lot of places that will be cheaper than Europe.
divbzero 4y ago
I am very sorry for your loss. I know there are efforts like Galleri [1] to detect cancer early using liquid biopsy [2] but I don’t know how effective they are at the moment. [1]: https://www.galleri.com/ https://www.galleri.com/ [2]: https://en.wikipedia.org/wiki/Liquid_biopsy https://en.wikipedia.org/wiki/Liquid_biopsy
zemvpferreira 4y ago
My condolences for your loss. My own father is a physican and very against any elective exams himself. His reasoning is that his colleagues are excessively interventive so the less medicine, the better. I wish your own father had been diagnosed in time but I can see the downside in doing too many exams as well. Doctors won't be able to keep their hands off treating any little lump they find, even if it's to your detriment. Case in point, surgery for back pain causes said pains to worsen at about the same rate as it improves them.

derekja 4y ago

sequencing continue to drop in price far faster than moore's law. https://www.genome.gov/about-genomics/fact-sheets/DNA-Sequencing-Costs-Data https://www.genome.gov/about-genomics/fact-sheets/DNA-Sequen... Here is the archive of the wired article, btw: https://web.archive.org/web/20221001173347/https://www.wired.com/story/the-era-of-fast-cheap-genome-sequencing-is-here/ https://web.archive.org/web/20221001173347/https://www.wired...

trention 4y ago
Not if you look at the period after 2015 only.
dekhn 4y ago
sequencing stopped dropping in real prices 5+ years ago.

peter303 4y ago

The second identifiable person to have his genome sequenced- James Watson- had around 30 known genome defects at the time. However most of them did not manifest as disease. That is a mystery.

FollowingTheDao 4y ago
SNP changes in genes are not usually mutations that cause "defects". Most usually they increase or decrease risk for a disorder based on more of a polygenic profile.
tejtm 4y ago
For small enough values of defect, it is par for the course.
dekhn 4y ago
No, it's only a mystery for people who have a truly simplified version of genetics and its relationship to disease phenotypes.

sometimeshuman 4y ago

Can someone recommend where one can order a full genetic sequence currently in the US and get the raw data ? It seemed highly-regulated when I last checked and required it be order through your doctor. The article touches on this a bit how the price gets into the thousands due to markups and counseling. Also how many full sequences would I have to get before the data is statistically sound. Since there is likely a base pair or two that will be different between the cells (maybe) and the sequencer hardware and process will likely have errors too. My gut tells me 10 to 20 but perhaps that seems like overkill since many base-pairs might be junk/repeats/irrelevant anyway and the raw data will have some confidence indication as well IIRC.

jghn 4y ago
You should be looking for 30x coverage. There are a number of companies that will do 30x WGS and give you the BAM file. I believe Nebula is one of them, but know there are other names out there.
- aroch 4y ago
  Nebula most likely ship your samples to China for processing and sequencing because it's so much cheaper. While that might not be a deal breaker for some, I would personally prefer my sample not be shipped elsewhere.
- FollowingTheDao 4y ago
  Stay away from Nebula. I had such a hassle with them and after six months of complaining with mostly no response I had to threaten then with a civil suit and they finally replied "your sample was no good, here is your money back."
chrisrohlf 4y ago
I asked this same question not long ago and the CEO of a well known company in this same research area reached out and recommended Veritas Genomics https://veritasgenetics.com/mygenome/ https://veritasgenetics.com/mygenome/ I’ve yet to finish the order process but I share this here because in my research of companies providing full sequencing services I never came across them and yet they seem far more reputable than others. Current SEO for these services turns up a lot of sketchy companies.

gavinray 4y ago

It's not common knowledge, but there's an algorithm called "genomic imputation" It can take incomplete data, like the kind you get from 23andMe, etc, and magically "fill in the gaps" to compute the majority of the rest of your genome There used to be a free service https://impute.me https://impute.me that would do this for you https://twitter.com/lassefolkersen/status/1547637086349537284?t=ewGfTYY6DywjCwrMRGbWSQ&s=19 https://twitter.com/lassefolkersen/status/154763708634953728... Search Google Scholar for "Lasse Folkersen", he is active in the field and has a lot of great papers about this

bilsbie 4y ago

I recall this being predicted something like 10 years ago. I wonder why a slowdown happened.

haihaibye 4y ago

Cheap* *To produce the raw data, and does not include analysis It's likely the bill for a consultant (senior doctor) to examine a patient, order a test then read and sign off on a report is already more than the chemistry + cpu + analysis already.

pazimzadeh 4y ago

It's good news but it's not as great as many people used to think. Having a particular gene doesn't mean that it's actually turned on. Your heart and your hair cells have the exact same genome, but do vastly different things. I am not excited as I used to be about the promise of cheap genome sequencing by itself. It needs to be paired with lots of other data for it to have any use, except for determining phylogeny/ancestry. Even more exciting than this would be cheap DNA synthesis, allowing us to build and test complicated genetic circuits in fewer steps. The current methods usually rely on stitching together existing sequences, rather than creating what you need from scratch. The price of DNA sequencing has fallen much faster than the price of DNA synthesis. Here is a comparison of the cost of DNA sequencing vs synthesis: http://www.synthesis.cc/synthesis/2016/03/on_dna_and_transistors http://www.synthesis.cc/synthesis/2016/03/on_dna_and_transis.... It would be good to find more recent data though.