Author Topic: A.I. uses "retrosynthesis" to copy patented drugs without infringing patents  (Read 391 times)

0 Members and 1 Guest are viewing this topic.

Offline dopamine

  • Subordinate Wasp
  • ***
  • Posts: 242
I thought this was pretty interesting, basically this company has developed AI that can reverse engineer drug patents in order to sidestep copyright infringement.  I bet with some tweaking it could also be purposed to sidestep drug laws and create new designer drugs / research chemicals.

Quote
Called Chematica, the software platform does something called “retrosynthesis,” similar to the kind of reverse engineering that takes place when an engineer dissects an existing product to see how it works. In the case of Chematica, this process is based on a deep knowledge of how chemical interactions take place. It has around 70,000 synthetic chemistry “rules” coded into its system, along with thousands of additional auxiliary rules prescribing when particular reactions occur and with which molecules they’re compatible. An algorithm then inspects the massive number of possible reaction sequences in order to find another way to the same finish line.
source: hxxps://www.digitaltrends.com/cool-tech/ai-develop-drug-without-trademark/

Offline gd

  • Larvae
  • *
  • Posts: 4
Quote
The rules are the "basic moves" that are used by graph-search algorithms to navigate enormous trees of synthetic possibilities in intelligent ways.

I just knew I wasn't barking up the wrong tree with my own endeavors (www.thevespiary.org/talk/index.php/topic,16556)...

Offline chip

  • Larvae
  • *
  • Posts: 34
  • a little bit more
    • dopetalk / d&u
i don't see how an alternate synthesis 'gets around' any such copyright infringement but i know nothing about the law.

i do, however, have a grasp on fair play.

surely it is moreso the product that bears the weight of copyright here, regardless of the route taken ?

mind you and having said that, i'm not sure what would happen if software synthed every possible molecular structure, so the method is definitely intellectual property.

Offline loft

  • Subordinate Wasp
  • ***
  • Posts: 208
Chematica was covered here before, awesome software, I which I had excess to it :D

Offline gd

  • Larvae
  • *
  • Posts: 4
Chematica was covered here before, awesome software, I which I had excess to it :D

I would love to play around with Chematica myself, if only I had the chance, but, these huge pieces of proprietary software that charge like $10,000 a license (or something like that) aren't always that much more sophisticated than what open source projects can do (with the right amount of dedication and contribution, at least).

It's like Wolfram / Mathematica / etc... Sure, super useful software if you can afford it (or can get a license by some other means), but totally replicable.

The awesome thing about computer science is that the investments are virtually non-existent. I've only recently gotten into chemistry and it's already costing me a small fortune (glassware, precursors, solvents, and the especially fancy shit like Soxhlet extractors, etc).
« Last Edit: February 02, 2019, 02:17:15 AM by gd »

Offline gd

  • Larvae
  • *
  • Posts: 4
i don't see how an alternate synthesis 'gets around' any such copyright infringement but i know nothing about the law.

i do, however, have a grasp on fair play.

surely it is moreso the product that bears the weight of copyright here, regardless of the route taken ?

mind you and having said that, i'm not sure what would happen if software synthed every possible molecular structure, so the method is definitely intellectual property.

Ok I'm most definitely not particularly knowledgeable when it comes to chemistry (or law, for that matter), but...

It seems to me like particular molecular configurations can't possibly be copyrighted since they're totally naturally occuring (even if synthesized in a lab). I mean, it feels like if someone were to copyright gasoline (goodbye 90% of the fuel industry).

I'm not speaking from a place of certainty or anything, but intuitively it just seems like it can't possibly be.

Even syntheses the though... Like, can someone copyright hydrocarbon combustion????

Some food for thought at the very least, that's for sure