4 ms·

Sequencing doesn't matter. It's getting the inbreds that matter. It's all in the article, but probably not so clear if you lack the background. Getting a bag o

by hibikir 11y ago

Sequencing doesn't matter. It's getting the inbreds that matter. It's all in the article, but probably not so clear if you lack the background. Getting a bag of GMO corn seed is not difficult. But reproducing said seed to the same level of quality is going to take you many years. It's a bit like trying to get a clone of yourself by having children with women: The kids will be a lot like you genetically, but it'll take you forever to make a clone. So how do agricultural companies do this? By building special inbred varieties that have he same genotype as phenotype. All the offspring of crossing any specific pair of inbreds will be identical, and those are the offspring that you can buy at the store. The inbreds, who are also worse performing in the field, not so much: They are seen as trade secrets, and the companies just use them for research and manufacturing. So it's not really about hiding the genetics of something you sell, or even hiding the genetics of something you use to create what you sell(which would still have some value, but still), but protecting the seeds of specific plants that are necessary to the manufacturing process. If you give some corn breeders an entire bag of hybrid corn, and tell them to produce 1000 bags just like it, they'll ask you to come back in a decade or two. If you give them a bag of each of the inbreds that were use to make it,they will have the 1000 bags in a season. That's what agricultural companies try to protect.

stephengillie 11y ago
You make these inbred parent seeds sound like private keys in encryption. You jealously guard them, you combine them to create offspring (public keys) which you freely distribute without worry, and lots of people are trying some arcane process to deduce your exact secret.
- btown 11y ago
  I think you mean public keys in the second sentence, but if so, that sounds like a very apt analogy.
  - stephengillie 11y ago
    Yes I did, and thank you.
- scarmig 11y ago
  Musing: Hmm, so let's play with Diffie-Hellplant. So you have two inbred genomes as two secrets, persisted as individual organisms. Alice and Bob negotiate on a public inbred genome that they pull from some kind of seed store (being careful to never re-use one they've used before!). Alice and Bob both create their own crossbreeds deterministically, from the secret organism and the public one. It's hard to get the private genomes from the two crosses, so they can send the seeds in the mail to each other. After that, I'm not sure you can get two shared secret clones from the crossbreeds and the opposite private inbreeds. (ETA: Pretty positive you can't, though if anyone has any ideas along these lines, I'm eager to hear!) Sounds like the start of some cypher-biopunk novel, though.
  - gknoy 11y ago
    Until this moment, I had not realized I would find such a book interesting.
    - scarmig 11y ago
      If you're interested in sci-fi where biotech trumps cybertech, The Windup Girl is a great read. Relevant to the article, too. It's set in a post-carbon dystopian Bangkok, where giant agribusiness monopolies dominate the world through control of the food supply, including offensive bioengineering to create mutant plagues that destroy native crops and introduce dependency. The main character is an agent of one of these agricorps, whose mission is espionage against Thailand's central seedbank.
      - DanBC 11y ago
        Also "Calorie Man". Seemingly available here: http://www.nightshadebooks.com/Downloads/WindupStories.pdf http://www.nightshadebooks.com/Downloads/WindupStories.pdf
        ethbro 11y ago
        "Calorie Man" reminded me a lot of "New Rose Hotel" by Gibson. Funny how the same story filters through different lenses as the years tick past.
- thaumasiotes 11y ago
  But in this case, deducing the exact secret is trivially easy; everyone knows how to do it.
  - stephengillie 11y ago
    Yes, which is why we're discussing Chinese people digging corn seeds out of American farms.
thaumasiotes 11y ago
> It's [...] probably not so clear if you lack the background. > So how do agricultural companies do this? By building special inbred varieties that have [t]he same genotype as phenotype. Are you sure you have "the background"? What do you mean by this? Genotypes and phenotypes can't be said to be "the same" in any dimension; they cover completely different ideas. Possible genotypes are things like "A allele" or "G allele", where phenotypes are things like "tall" or "slimy". If you mean that the company strains are all homozygous at certain locations of interest, and the seed they sell is heterozygous, that just means that you'll lose a certain amount of seed to quality issues when you breed your seeds directly. It also means you can trivially recreate the company strains by lifting chromosomes out of non-gamete cells. (Granted, I have no idea where the technology is for this -- it's cloning in the conventional "I have a biology lab" sense.) If you mean that the company strains actually have just the one chromosome N times over for each chromosome, that's hard to believe. An animal would almost definitely immediately die with genetic makeup like that. But, again, genetic engineering could still do this. Also, there's no way the genetic information of these company strains could be considered "trade secrets", since it's all freely released to the public in the hybrid strains.
- sampo 11y ago
  > can trivially recreate the company strains by lifting chromosomes out of non-gamete cells. That is an interesting idea. Corn has 10 pairs of chromosomes. I wonder if you can tell for each pair, which one came from which parent? But then again, it shouldn't matter. As long as a set of 10, and its complement, both produce viable individuals, their offspring should be identical to the desired hybrid. Now we just must not tell this to the Chinese...
  - thaumasiotes 11y ago
    > Corn has 10 pairs of chromosomes. I wonder if you can tell for each pair, which one came from which parent? For humans the chromosomes are marked in this way. I don't know about corn.
    - sampo 11y ago
DanBC 11y ago
So, why do the companies aggressively prevent seed keeping and sowing? Why do farmers have to buy seed each year? If the harvested seed isn't the same as the sold seed what's the problem for the seed company? (You can probably tell that my information comes from mostly anti GMO sources. But I'm not particularly anti GMO)

by hibikir 11y ago

stephengillie 11y ago

You make these inbred parent seeds sound like private keys in encryption. You jealously guard them, you combine them to create offspring (public keys) which you freely distribute without worry, and lots of people are trying some arcane process to deduce your exact secret.

btown 11y ago
I think you mean public keys in the second sentence, but if so, that sounds like a very apt analogy.
- stephengillie 11y ago
  Yes I did, and thank you.
scarmig 11y ago
Musing: Hmm, so let's play with Diffie-Hellplant. So you have two inbred genomes as two secrets, persisted as individual organisms. Alice and Bob negotiate on a public inbred genome that they pull from some kind of seed store (being careful to never re-use one they've used before!). Alice and Bob both create their own crossbreeds deterministically, from the secret organism and the public one. It's hard to get the private genomes from the two crosses, so they can send the seeds in the mail to each other. After that, I'm not sure you can get two shared secret clones from the crossbreeds and the opposite private inbreeds. (ETA: Pretty positive you can't, though if anyone has any ideas along these lines, I'm eager to hear!) Sounds like the start of some cypher-biopunk novel, though.
- gknoy 11y ago
  Until this moment, I had not realized I would find such a book interesting.
  - scarmig 11y ago
    If you're interested in sci-fi where biotech trumps cybertech, The Windup Girl is a great read. Relevant to the article, too. It's set in a post-carbon dystopian Bangkok, where giant agribusiness monopolies dominate the world through control of the food supply, including offensive bioengineering to create mutant plagues that destroy native crops and introduce dependency. The main character is an agent of one of these agricorps, whose mission is espionage against Thailand's central seedbank.
    - DanBC 11y ago
      Also "Calorie Man". Seemingly available here: http://www.nightshadebooks.com/Downloads/WindupStories.pdf http://www.nightshadebooks.com/Downloads/WindupStories.pdf
      - ethbro 11y ago
        "Calorie Man" reminded me a lot of "New Rose Hotel" by Gibson. Funny how the same story filters through different lenses as the years tick past.
thaumasiotes 11y ago
But in this case, deducing the exact secret is trivially easy; everyone knows how to do it.
- stephengillie 11y ago
  Yes, which is why we're discussing Chinese people digging corn seeds out of American farms.

thaumasiotes 11y ago

> It's [...] probably not so clear if you lack the background. > So how do agricultural companies do this? By building special inbred varieties that have [t]he same genotype as phenotype. Are you sure you have "the background"? What do you mean by this? Genotypes and phenotypes can't be said to be "the same" in any dimension; they cover completely different ideas. Possible genotypes are things like "A allele" or "G allele", where phenotypes are things like "tall" or "slimy". If you mean that the company strains are all homozygous at certain locations of interest, and the seed they sell is heterozygous, that just means that you'll lose a certain amount of seed to quality issues when you breed your seeds directly. It also means you can trivially recreate the company strains by lifting chromosomes out of non-gamete cells. (Granted, I have no idea where the technology is for this -- it's cloning in the conventional "I have a biology lab" sense.) If you mean that the company strains actually have just the one chromosome N times over for each chromosome, that's hard to believe. An animal would almost definitely immediately die with genetic makeup like that. But, again, genetic engineering could still do this. Also, there's no way the genetic information of these company strains could be considered "trade secrets", since it's all freely released to the public in the hybrid strains.

sampo 11y ago
> can trivially recreate the company strains by lifting chromosomes out of non-gamete cells. That is an interesting idea. Corn has 10 pairs of chromosomes. I wonder if you can tell for each pair, which one came from which parent? But then again, it shouldn't matter. As long as a set of 10, and its complement, both produce viable individuals, their offspring should be identical to the desired hybrid. Now we just must not tell this to the Chinese...
- thaumasiotes 11y ago
  > Corn has 10 pairs of chromosomes. I wonder if you can tell for each pair, which one came from which parent? For humans the chromosomes are marked in this way. I don't know about corn.
  - sampo 11y ago

DanBC 11y ago

So, why do the companies aggressively prevent seed keeping and sowing? Why do farmers have to buy seed each year? If the harvested seed isn't the same as the sold seed what's the problem for the seed company? (You can probably tell that my information comes from mostly anti GMO sources. But I'm not particularly anti GMO)