01-15-2020, 08:47 PM
Link to the original paper:
A scalable pipeline for designing reconfigurable organisms, Kriegman et al (2020)
https://www.pnas.org/content/early/2020/...1910837117
It’s a great piece of research. The researchers developed a virtual physics model of the basic features of specific frog cells (biomechanical properties, behaviour in response to chemical cues etc). They then ran evolutionary algorithms for a long time on some serious computing hardware. Assemblages of cells would be randomly created and tested in the environment against a number of tasks. The best designs would be kept and reproduced for further testing. Eventually they had some designs of cell assemblages that could coordinate and move.
They then tested these designs in vitro. Using microsurgery they grafted cells together into the arrangements suggested. Some of them performed as expected: able to move, carry objects, push objects etc. What’s really cool about this work is that it demonstrates how cell and tissue behaviour is strongly determined by environmental cues. The pop-sci/laymen understanding is that DNA is a blue print or an instruction tape. But biological behaviour is an emergent property of the environment. So even though the DNA is completely unaltered by simply changing tissue arrangement in specific ways novel tissue behaviour can be gained.
In effect these are a sort of Frankenstein micro-organism. Both the methodology and the result are equally promising for future developments of micro-biobots.
A scalable pipeline for designing reconfigurable organisms, Kriegman et al (2020)
https://www.pnas.org/content/early/2020/...1910837117
It’s a great piece of research. The researchers developed a virtual physics model of the basic features of specific frog cells (biomechanical properties, behaviour in response to chemical cues etc). They then ran evolutionary algorithms for a long time on some serious computing hardware. Assemblages of cells would be randomly created and tested in the environment against a number of tasks. The best designs would be kept and reproduced for further testing. Eventually they had some designs of cell assemblages that could coordinate and move.
They then tested these designs in vitro. Using microsurgery they grafted cells together into the arrangements suggested. Some of them performed as expected: able to move, carry objects, push objects etc. What’s really cool about this work is that it demonstrates how cell and tissue behaviour is strongly determined by environmental cues. The pop-sci/laymen understanding is that DNA is a blue print or an instruction tape. But biological behaviour is an emergent property of the environment. So even though the DNA is completely unaltered by simply changing tissue arrangement in specific ways novel tissue behaviour can be gained.
In effect these are a sort of Frankenstein micro-organism. Both the methodology and the result are equally promising for future developments of micro-biobots.
OA Wish list:
- DNI
- Internal medical system
- A dormbot, because domestic chores suck!