ABACUS Project has achieved a breakthrough result in Biocomputation
- Published in Proceedings of the National Academy of Sciences - 113, 2591-2596 (2016) -
Dan V. Nicolau Jr., Mercy Lard, Till Korten, Falco C. M. J. M. van Delft, Malin Persson, Elina Bengtsson, Alf Månsson, Stefan Diez, Heiner Linke, and Dan V. Nicolau
The most important achievement of the ABACUS Project in the field of Biocomputation is the outstanding proof of the concept for parallel computing based on designed networks explored by self-propelled biological agents. This could well be the early stage of an energy-efficient super-biocomputer using proteins, with superior performances than those of the current quantum computers.
A recent scientific article resulted from ABACUS Project funded by the European Commission, has been published on February 2016 in the Proceedings of the National Academy of Sciences of the United States - PNAS - http://www.pnas.org/content/early/2016/02/17/1510825113.abstract?sid=e3f3d0e6-d901-429a-99eb-e90f491686d3. This result is in many respects the culmination of the last decade of collaborative work between the members of ABACUS Project Consortium.
The recent published scientific article describes an early version of how such a "living computer” might look and function, using the ATP (Adenosine Triphosphate) molecules that power muscle and brain signalling, among many other critical biological processes. The research is not a software simulation, but an experimental proof of the concept. Firstly, a difficult mathematical problem was converted into a maze-like nanostructure. The ATP-powered proteins, which cover the walls of the network channels, propel the biomolecular agents, which randomly explore this purposefully-designed maze, finding all the solutions of the mathematical problem at the end of their journey. The hallmark of living things – the parallelism in solving problems, has been mastered by design and by using many times less energy than the equivalent electronic computer would require. Further work could demonstrate that such "bio-computers" can perform calculations that normal computers cannot.
Strings of proteins of different lengths travel around the nano-network that powers a new model biological supercomputer.
Image credits ABACUS Project, Till Korten, Dresden University
We all have the sense that, on one hand living beings are constantly ‘solving’ problems; and on the other hand, they do so in a very different way than human-designed computers. It happens that this ‘methodology’ is extraordinarily powerful, and it is also remarkably efficient at solving very complex problems. So how are the two ways different? Nature computes in a different way than electronic computers, proceeding in a massively parallel fashion rather than sequentially. Humans, for example, can think, breathe, regulate their heart function, love or be angry, all the same time, even in our sleep. So, while Nature appears to be the ultimate master of simultaneity or synchronicity, human-designed computers excel at one-at-a-time, repetitive tasks. So, what can Life figure out that we cannot?
In some ways, this dilemma is encapsulated in the famous so-called "P vs NP problem" (http://www.claymath.org/millennium-problems/p-vs-np-problem). For instance, it is essentially impossible to organize housing accommodations for four hundred student applicants, with space limited to one hundred, and with a list of pairs of incompatible students – the total number of ways of choosing one hundred students from the four hundred applicants is greater than the number of atoms in the known universe! On the engineering front, can we build computers that use living "components" and the same engineering principles as living things? And can these computers solve these "NP-complete problems" mentioned above that appear not to have efficient conventional algorithms? More pragmatically, if we could harness components of living things or simple microorganisms to do 'natural' problem solving, what kind of new mathematics will emerge?
In mathematics, we define algorithms as step-by-step recipes for solving mathematical problems, and the main reason why computers have revolutionised our world over the past half-century is that we can often find "quick" algorithms for many practical problems of interest: GPS navigation in a car, number crunching in spreadsheets, or designing bridges. Not always: we have known since the 1960s that some problems appear not to have such quick recipes (these are the "NP-complete" problems). They include finding faces in pictures, cracking cryptographic codes and, most fascinatingly, the ability of being creative, turning thoughts into reality.
Whether these problems do in fact have clever recipes, and humans are not clever enough to find them, or whether it is a profound property of Nature that these problems cannot be solved using electronic computers, is the greatest mystery of computer science and perhaps of all mathematics. Beyond the technicalities, this question about building biocomputers is pregnant with philosophical implications, because if it were the case that all problems have efficient computer algorithms, then creativity being only a human capability would be just an illusion. If such algorithms could be find, anyone who can appreciate a symphony would be able to compose like Mozart, everyone who can understand a mathematical proof would be as much of a mathematical genius as Einstein. That doesn't fit yet with our experiential knowledge of the world, but we do not yet know for sure — and we are unlikely to find out soon — the answer to this mystery within a mystery.