4 ms·

Molecular Programming Project

82 points by MichaelAO 11y ago

dluan 11y ago
This is almost everyone involved in the growing field of synthetic and molecular biology tools. Kind of neat to see and recognize these names, a bit surreal for me personally because this is the group and area I would've ended up in! Such fond memories of getting gro to run properly...
_pmf_ 11y ago
> Computer science and engineering has mastered complexity for electronic computation That's under debate.
- crimsonalucard 11y ago
  I wouldn't even say it's under debate. We clearly haven't mastered it.
- femto 11y ago
  Rather then mastering complexity, I'd say engineering, and in particular electronic engineering, has mastered prediction. It was the rise of accurate simulation tools, such as SPICE, that allowed the construction of reliable integrated circuits and the computing revolution. It strikes me that the nano/molecular technology revolution depends on the wide availability of accurate simulators. (Please do say if there already a molecular equivalent of SPICE available!)
  - crimsonalucard 11y ago
    Yes there is! But the complexities of the real world make it so that they are significantly more complex than pspice. They are literal physical simulations in 3D space. https://www.gpugrid.net/science.php https://www.gpugrid.net/science.php
- wyager 11y ago
  Computers are by far the most complex machines humans have ever made (in terms of number of components, information required to describe their construction, etc.). By that standard, electrical/computer engineering routinely manages more complexity than any other field.
  - crimsonalucard 11y ago
    Computer engineering abstracts complexity into it's own little perfect universe. CE may seem more complex because this, but it is actually LESS complex as a result. Due to the lower complexity of the virtual universe, humans can build structures of seemingly greater sophistication, but the sophistication is an illusion as the built system is actually simpler due to the fact that the artificial universe in which it is created is in itself, simpler. Additionally, due to the often virtual nature of the application these "virtual" systems can and are often deployed with many bugs and components that are broken. An example of a field with greater (or matching) complexity than computer engineering is aerospace engineering. The real world effects on a physical system, say the billion dollar F-22 or boeing 787, make it a wildly more complex system than say facebook.com. Realities and unknowns of the physical world heavily influence the performance of these systems and as a result the complexities associated with the physical world become reflected in the system designed to operate within it. Additionally, these real-world systems demand near perfection in execution due to the non-virtual (often dangerous) consequences that these systems may pose as a result of the system's existence in the real world.
versteegen 11y ago
That's a lot of different research directions. But I didn't see P systems mentioned. Maybe because P systems are a model of computation studied by computer scientists that assumably isn't currently implementable? There are many variants, but the basic model is of set a compartments (e.g. cells) which contain signalling chemicals which can be synthesised, destroyed, or passed between adjacent compartments according to rules. A number of P system algorithms have been developed such as leader election and graph diameter calculation (of the adjacency graph). Seems very difficult to program. https://en.wikipedia.org/wiki/P_system https://en.wikipedia.org/wiki/P_system
- fenollp 11y ago
  Do you have a link to a leader election algorithm in P sytems? My Google-fu returns nothing…
hellbanner 11y ago
Gamification Mechanic Wiki, for those interested in learning more about the process or building your own. https://badgeville.com/wiki/Game_Mechanics https://badgeville.com/wiki/Game_Mechanics
agumonkey 11y ago
Reminded me of the "The Chemical Abstract Machine" http://www.lix.polytechnique.fr/~fvalenci/papers/cham.pdf http://www.lix.polytechnique.fr/~fvalenci/papers/cham.pdf
arca_vorago 11y ago
This reminds me I need to double check the end date of my non-compete in bioinformatics. I learned so much while I was doing it and I still find the field to be ripe for startups like this. I wish them all the best, and I have always been of the mind that we won't ever get good brain emulation any other way than a combination of traditional silicon and biocomputing.
stillsut 11y ago
I think what's missing is "F5". That's what gave us the web: a bunch a yahoos (without knowing the formal rules) kept thrashing on that key until they got something to work. My 2cents: we need more yahoos and we need an quick and reliable F5 - a test to see if the hypothesis is verified.

82 points by MichaelAO 11y ago

dluan 11y ago

This is almost everyone involved in the growing field of synthetic and molecular biology tools. Kind of neat to see and recognize these names, a bit surreal for me personally because this is the group and area I would've ended up in! Such fond memories of getting gro to run properly...

_pmf_ 11y ago

> Computer science and engineering has mastered complexity for electronic computation That's under debate.

crimsonalucard 11y ago
I wouldn't even say it's under debate. We clearly haven't mastered it.
femto 11y ago
Rather then mastering complexity, I'd say engineering, and in particular electronic engineering, has mastered prediction. It was the rise of accurate simulation tools, such as SPICE, that allowed the construction of reliable integrated circuits and the computing revolution. It strikes me that the nano/molecular technology revolution depends on the wide availability of accurate simulators. (Please do say if there already a molecular equivalent of SPICE available!)
- crimsonalucard 11y ago
  Yes there is! But the complexities of the real world make it so that they are significantly more complex than pspice. They are literal physical simulations in 3D space. https://www.gpugrid.net/science.php https://www.gpugrid.net/science.php
wyager 11y ago
Computers are by far the most complex machines humans have ever made (in terms of number of components, information required to describe their construction, etc.). By that standard, electrical/computer engineering routinely manages more complexity than any other field.
- crimsonalucard 11y ago
  Computer engineering abstracts complexity into it's own little perfect universe. CE may seem more complex because this, but it is actually LESS complex as a result. Due to the lower complexity of the virtual universe, humans can build structures of seemingly greater sophistication, but the sophistication is an illusion as the built system is actually simpler due to the fact that the artificial universe in which it is created is in itself, simpler. Additionally, due to the often virtual nature of the application these "virtual" systems can and are often deployed with many bugs and components that are broken. An example of a field with greater (or matching) complexity than computer engineering is aerospace engineering. The real world effects on a physical system, say the billion dollar F-22 or boeing 787, make it a wildly more complex system than say facebook.com. Realities and unknowns of the physical world heavily influence the performance of these systems and as a result the complexities associated with the physical world become reflected in the system designed to operate within it. Additionally, these real-world systems demand near perfection in execution due to the non-virtual (often dangerous) consequences that these systems may pose as a result of the system's existence in the real world.

versteegen 11y ago

That's a lot of different research directions. But I didn't see P systems mentioned. Maybe because P systems are a model of computation studied by computer scientists that assumably isn't currently implementable? There are many variants, but the basic model is of set a compartments (e.g. cells) which contain signalling chemicals which can be synthesised, destroyed, or passed between adjacent compartments according to rules. A number of P system algorithms have been developed such as leader election and graph diameter calculation (of the adjacency graph). Seems very difficult to program. https://en.wikipedia.org/wiki/P_system https://en.wikipedia.org/wiki/P_system

fenollp 11y ago
Do you have a link to a leader election algorithm in P sytems? My Google-fu returns nothing…

hellbanner 11y ago

Gamification Mechanic Wiki, for those interested in learning more about the process or building your own. https://badgeville.com/wiki/Game_Mechanics https://badgeville.com/wiki/Game_Mechanics

agumonkey 11y ago

Reminded me of the "The Chemical Abstract Machine" http://www.lix.polytechnique.fr/~fvalenci/papers/cham.pdf http://www.lix.polytechnique.fr/~fvalenci/papers/cham.pdf

arca_vorago 11y ago

This reminds me I need to double check the end date of my non-compete in bioinformatics. I learned so much while I was doing it and I still find the field to be ripe for startups like this. I wish them all the best, and I have always been of the mind that we won't ever get good brain emulation any other way than a combination of traditional silicon and biocomputing.

stillsut 11y ago

I think what's missing is "F5". That's what gave us the web: a bunch a yahoos (without knowing the formal rules) kept thrashing on that key until they got something to work. My 2cents: we need more yahoos and we need an quick and reliable F5 - a test to see if the hypothesis is verified.