FIRST OPEN-INNOVATION BIO4COMP AWARD (5 000 Euro)

The first open-innovation Bio4Comp Award was for novel ideas that conceptually or technically advance the field of parallel computing with biological agents in a substantial manner. The specific challenge was to find a way to tag cytoskeletal filaments. The deadline was on 15th January 2019. Bio4Comp would like to thank all participants for their creative ideas.

Winning idea: Colorimetric path tagging of filaments using DNA-based metafluorophores

By Prof. Adam P. Micolich
University of New South Wales, Sydney, Australia

The full description of the idea has been published on arXiv:

https://arxiv.org/abs/1909.06024

Challenge: Encoding and decoding of information from cytoskeletal filaments for parallel computing

For a number of mathematical problems to be solved by NBC approaches, it will be crucial to efficiently encode information into the cytoskeletal filaments themselves. The latter will increase the flexibility of network design and open the road to solving a variety of different complex problems.

Here, we invite proposals that address the issues of physical and/or (bio)chemical information-encoding and information-readout, e.g. to obtain information about the path a filament has taken through a network.

Successful information encoding requires the ability to write information into (“tag”) filaments in regions of a few micrometers size, and to read out that information, at the single-filament level, at detection sites that are located hundreds of micrometers or more away from the writing regions.

The writing mechanism should allow writing of at least 10, ideally up to 100s of different tags at a minimum of 10, ideally 100s of different sites.

The information must be written into individual filaments or molecules carried by those filaments.

The read out mechanism must allow the decoding of the recorded information from at least thousands, ideally millions or even billions of individual filaments.

The combinatorial nature of many important mathematical problems, including so called NP-complete problems (nondeterministic-polynomial-time-complete  problems), places a severe limitation on the problem size that can be solved with conventional, sequentially operating electronic computers. While a number of parallel-computation approaches, such as DNA computation, quantum computation, and microfluidics-based computation, have been reported in the past, these approaches have so far not proven to be scalable and practical from a fabrication and operational perspective.

One alternative is network-based computation (NBC) where a given combinatorial problem is encoded into a graphical, modular network that is embedded in a nanofabricated planar device. Exploring the network in a parallel fashion using a large number of independent agents (e.g. cytoskeletal filaments propelled by molecular motor proteins) then solves the mathematical problem.

This approach uses orders of magnitude less energy than conventional computers, thus addressing issues related to power consumption and heat dissipation (see: Nicolau et al., PNAS, 113(10), pp. 2591-2596, 2016).