Researchers report a method for designing simple organisms with custom functions, in which a computer was used to generate candidate organisms out of specified biological building blocks based on an evolutionary algorithm; using a combination of contractile and passive living tissues, the authors built high-performing candidates that exhibited predicted behaviors and capabilities, and the authors suggest that such a method could be generalized to design living machines for various applications, such as smart vehicles for drug delivery and internal surgery.
###
Article #19-10837: "A scalable pipeline for designing reconfigurable organisms," by Sam Kriegman, Douglas Blackiston, Michael Levin, and Josh Bongard.
MEDIA CONTACT: Josh Bongard, University of Vermont, Burlington, VT; tel: 802-656-4665, 802-578-4445; e-mail: < josh.bongard@uvm.edu >; Michael Levin, Tufts University, Medford, MA; e-mail: < michael.levin@tufts.edu >
Proceedings of the National Academy of Sciences