The future of MEBC: panel discussion

BioSystems

35 (1995) 233-237

The future of MEBC: panel discussion* G. Bin6 Central Research Institute of Chemistry, Hungarian Academy of Sciences, H-1525 Budapest 114, P.O. Box 17, Hungary

Abstract The expected developments in the not too distant future (5 10 years) of molecular electronics and biocomputing (MEBC) are discussed. In the short-term, the study of very specific basic phenomena is expected (e.g. conducting polfmers, strange electronic states of insulating polymers, bacteriorhodopsin (BR), arrays of molecules, self-organization of biomaterials, very specific biological systems, quantum coherence in cytoskeletal microtubules, optoelectronic information storage, associative memories, pattern recognition, hierarchical nature of biological information). New application fields outside the range of conventional technology (e.g. randomized algorithms, optoelectronic devices, chemical and biosensors, as well as a certain extent of commercialisation) have also been predicted. In the long-term, the study and solution of much deeper (sometimes scientific fiction-like) problems were foreseen, such as the self-organization of biomaterials, artificial self-reproduction, implementation of artificial cell dynamic control structures based on molecular devices for medical and environmental applications and the construction of neuronal computers as aids to the human brain. Keywords:

processing;

Single molecular Biocomputing

circuits;

Self-organization;

Optoelectronics;

1. Introduction

devices/electronics;

tion in the responses. Interesting concise answers will be presented

Near the termination of this seminal meeting, in its rather excited atmosphere, the invited members of the closing panel were asked to fill in a transparency form with the text: ‘The title of my (our) lecture (1) in MEBC 1998:..., Author(s):...; and (2) in MEBC 2003:..., Author(s):...‘. They were also asked to answer the question ‘Why?’ and to answer a question from the audience. Both the great goals of molecular electronics in the distant future and the Smaller ones, the immediate problems of molecular engineering, received significant atten* Based on the final panel discussion at MEBC 93. 0303-2647/95/$09.50

Molecular

0 1995 Elsevier

SSDI 0303-2647(94)01521-S

Science

Ireland

Ltd. All rights

2. The future -

Information

thoughts below.

and

what do we expect?

2.1. From molecular devices to neuronal computers (Girjesh Govil) Please permit me to quote Professor Govil here without any change. ‘(1) in MEBC 1998, Molecular devices: passive devices (LCD); conducting wires/polymers; chemical and biosensors; optoelectronic devices/storage; pattern recognition; commercialisation. (2) in MEBC 2003, Neuronal computers as aids to human brain: sensors (smell, taste, toxicity); extension of visual range; health reserved

234

G. Biczd / BioSystems

hazard alarms; learning, memory, thought processes. In view of the fact that molecular electronics is an emerging area of science and technology where new developments are taking place at a very rapid rate and the fact that new scientists are entering these areas, it would not be correct to make any guess on the names of the scientists who will be the keynote lecturers in MEBC 1998 and MEBC 2003 .’ 2.2. Optical information processing using molecular arrays (Michael C. Petty) In the short-term (MEBC 1998, authors: M.C. Petty et al.), Professor Petty thought that the optical properties of organics would be important and suggested that useful devices may emerge from detailed studies of the interaction of arrays of dye molecules, biological systems and guided surface EM modes. These works will lead to design rules for macromolecules exhibiting simple processing functions. In the long-term (MEBC 2003, A macromolecular processing unit, M.C. Petty et al.), synthetic chemistry and biological aspects will become increasingly important. Why? This is so because the materials that we are working on today are unlikely to be those exploited in the next century. For example, work on LB films may just provide design rules to enable chemists to synthesize molecules containing a number of functional groups (Stoddart, 1993). Finally, Professor Petty added the following two points: (1) molecular electronic devices already exist. It is therefore not necessary to ask the question ‘When will organic devices become a reality?’ Examples are liquid crystal displays (it is doubtful whether these would have been predicted 30 years ago), photoreceptors for xerography and plastics (e.g. PVDF) for electret microphores and pyroelectric detectors; and (2) the definition of molecular electronics varies from country to country. Biological aspects seem to dominate in Japan and the USA, whereas the exploitation of the macroscopic properties of organic materials is probably more important in Europe.

3.5 (1995) 233-237

2.3. From biomolecular self-organization to artljicial self-reproduction (Norbert A. Hampp)

Very important problems have been pointed out not merely with the titles but also with the second authors. (1) in MEBC 1998, Self-organization of biomaterials, authors: N. Hampp and S.O.M.E. Luck and (2) in MEBC 2003, Self-reproducing artificial biosystems, authors: N. Hampp and B.I.O. Network. These are true projects for the coming 5-10 or more years, indeed, having a sound basis in his recent researches (Hampp, 1994a, b). Furthermore, the size and complexity of the tasks really require very broad team activities within extended bionetworks. We also have to follow his warning ‘Be careful with term biocomputing. Nobody knows a model how the brain works...’ For Klaus-Peter Zauner’s question ‘What are the possible applications of BR in neuronal network architectures (concerning optical implementations of technical neural networks)?, Dr Hampp pointed out that there is no intelligence in the BR and the applications so far are using ensembles of molecules and not single molecules. In clear concordance with the above titles, in the case of BR also, ensembles will certainly be applied as real-time holographic pattern recognition (Hampp, 1994a, b). 2.4. Not to compete with conventional technology

-

rather to $11 new niches (Klaus-Peter Zauner)

According to his short summary, molecular computing conveys the potential to access new application fields outside the range of conventional technology and in the foreseeable future, molecular computing will more likely fill these niches than compete with current technology. His first title, Non-deterministic parallel algorithms as a niche for molecular computers (both suggested without authors), has reflected the goal of transforming theoretical concepts into real architectures that conventional computers can only approximate. In his opinion, the class of problems that is non-computable for deterministic algorithms but tractable in a randomized setting is an example. The access to uncorrelated (i.e. not pseudo-random) information is a bottleneck for a conventional computer. This restriction becomes severe in a parallel computing environment. View-

G. Bid

/ BioSystems

35 (1995) 233-237

235

ing uncorrelated (random) information as a resource comparable to (memory) space and (execution) time, Zauner predicted that molecular devices will turn out to be superior for the implementation of algorithms with high random complexity (Prigogine, 1989; Traub and Wozniakowski, 1991). For 2003, the title has been given as: On the implementation of artificial control structures in E. coli for medical and environmental applications. This title was intended to point out another niche for molecular computing devices. Today’s bioengineering allows one to manipulate the molecular structure of cells. The field of metabolic engineering attempts to alter the enzymatic reaction chains in order to optimize the yield of a desired product (e.g. insulin produced by a modified microorganism). With the growing understanding of the natural control of intracellular mechanisms, the implementation of dynamic control structures based on molecular devices is conceivable Zauner concluded (quoting Cameron and Tong, 1993).

key to establishing a form of metabolic molecular machine would be equivalent to figuring out or fabricating a molecular process which can hold on to thermal agitations for much longer than the thermal relaxation time.

2.5. Metabolic molecular electronics (Koichiro Matsuno)

2.7. From the hierarchy of biological information

Matsuno has proposed this title for 1998, which he hopes will be authored by himself and his company. A basic reason for this title is to emphasize the significance of the capacity of the molecular machine for obtaining its own ‘food’ by itself. Thermal fluctuations modulated by the underlying molecular machine can exert an extractable ability to perform work. Modulated thermal fluctuations can be a most versatile and flexible source of work capacity. How could one fabricate such machines at all? This seems to be a major challenge for the year 2003. For the moderator’s question ‘How do you imagine the above basic energetic process, say, in a metabolic switch?, Professor Matsuno cited the actomyosin complex as an example. The sliding movement of an actin filament relative to myosin heads could be initiated by thermal fluctuations, but the holding time of those fluctuations within the system must be much larger than the available thermal relaxation time. This process of holding thermal agitations within a limited region requires a certain energy source to ‘feed upon’. In essence, a

This discussion appeared as a spontaneous contribution from the audience. (1) In MEBC 1998, Molecular computing and the hierarchical nature of biological information. Authors: a series of articles authored by the most active researchers in this field. As Dr Marijuan stated, molecular computing, as a new interdisciplinary field that utilizes biological molecules as fundamental tools for computation and information processing, can make substantial contributions to elucidate unanswered questions related to information flow in biological systems. Actually, other fields such as artificial intelligence and artificial life have skimmed the very basic problems at the cellular level. How can ensembles of molecules be organized in order to configure a basic ‘computing’ (information processing) entity? What is the hardware-software relationship for these molecules? What is the informational role of water (necessary to establish a host of parallel integrative biophysical and biochemical processes within the cell) and what is the significance of the protein synthesis/ degradation phenomena (both fundamental tools

2.6. Cytoskeletal microtubules: quantum coherence, devices, consciousness (Stuart R. Hamerofi

up to ethics

This is another wonderful new extended bio-scientific perspective! For a better understanding, a reading of the author’s latest beautiful book (Koruga et al. 1993) is suggested. Then it will be easier to understand the two projected titles. (1) in MEBC 1998, Quantum coherence in cytoskeletal microtubules: interface to nanoscale squid arrays. Authors: Hameroff, Dayhoff, Koruga, Penrose, Lahoz-Beltra and Samsonovich. (2) in MEBC 2003, Consciousness in cytoskeletal microtubule devices: ethical considerations. Author(s): Hameroff, Koruga, Dayhoff, Penrose, Lahoz-Beltra and Samsonovich.

to the system of sciences (Pedro C. Marijuan)

236

G. Biczd / BioSystems

in order to regulate the ‘population’ of acting enzymes and proteins)? Coherent modeling and formalization of these (an ‘Artificial Cell Project’) might represent a new starting point to chart the bidirectional flow of information and problemsolving activities that characterize multicellular organisms. The differentiated cell types represent topologically distributed sources of problems and solutions for every organism; without a rigorous ‘informational perspective’ of the cell itself, we won’t be able to go beyond the pragmatic approaches (e.g. physiology, medicine) and the merely syntactic ones (artificial intelligence). (2) In MEBC 2003, Information science and the system of the sciences. Authors: a multidisciplinary team involving philosophers, mathematicians, theoretical physicists and biologists, sociologists and molecular computing scientists. Beyond the informational characterization of, let us say, the ‘society’ of enzymes, the study can be extended to a series of information-based ‘societies’ of growing complexity: the ‘society of neurons’ and the ‘society of individuals’. The informational processes within these new entities constitute, at an abstract level, genuine ‘replicas’ of the problems and solutions already confronted/ discovered by the cellular system. The information processing in these entities continues to be a topologically distributed one, where members of a society become societies at the lower level, and the flow of information has to become ‘plastic’ between ‘social’ levels, fanning in and out at every interface, emerging and disappearing. Living cells endow organisms and nervous systems with very special properties; information processing emerges in a massive vertical scale (Conrad, 1990). Although it seems too speculative, perhaps our body of basic ‘horizontal’ disciplines (mathematics, physics, chemistry, biology, psychology and the social sciences) should be reconsidered in order to accommodate a new ‘vertical science’ devoted to information, instances of which would be present in today’s computer science and molecular biology (and particularly in their interface, molecular computing) and also in different areas of physics, biology and neurosciences. I think this spontaneous contribution offers an excellent foresight of a newly developing field dealing with the large-

35 (1995) 233-237

scale (‘vertical’ and ‘horizontal’) information processing through multi-superposed bio-(sub)structures and societies. Note that Hameroff s picture (above) inherently suggests the importance of both ‘vertical’ and ‘horizontal’ information flows in multi-superposed structures, to a real understanding of brain functioning. 2.8. From a single molecular circuit to a IO-year biocomputing project to help human gene corrections (G&a Biczb)

My own titles have represented desires or longings rather than expectations; although the aims involved seem to be realistic under optimal circumstances, I know that my actual possibilities are very far from those. The first title planned in 5 years’ time, The first single molecular circuit based on strange electronic states (SES) has been functioning for 1 month (authors: many many well known names in MEBC, from A to Z including myself) has been based on the considerations presented in my lectures (Biczo, 1994a, b). The tools - scanning tunnelling microscope (STM) and manipulator, Langmuir-Blodgett means and quantum chemical computer programmes - are already available to study the SES of relatively small polymers and, in case of success, to build up such a circuit (naturally, with extensive cooperation and sooner rather than later if the participants of this panel discussion would also collaborate with me). The second title, A lo-year project to develop molecular and biocomputers to help in the correction of human genes (CHG), was intended to point out a very serious problem of the human race. Today’s medicine, with the humanistic efforts to cure the most terrible diseases, is going to destroy the healthy genetic composition of the mankind. To deal with this problem and to proceed forward to a more promising future, genetic reconstruction and artificial gene developments are required. These are extremely difficult tasks that will need to use molecular and biocomputers as the most powerful tools. Concerning CHG, Jeff Paffmann addressed a question and a comment to me ‘Do we really want to ‘correct’ human genes? What we may think is a correction may eliminate an evolutionary avenue for the future. I can see the case for

G. Bid

/ BioSystems

gene correction to cure a cancer patient, but what happens if this process is used to change the attributes in an undeveloped egg? Given uncontrolled use we could guide our evolution into stagnation’. Now I think that dangers and advantages always come together and we always have to direct these ‘genetic cures’ with the highest possible care, both for individuals and for the whole of mankind. Because of the almost infinite number of realizable genetic variations, I believe that we could not guide our evolution just into stagnation. On the contrary, new pathways are more likely to open for evolution. For example, and a little bit symbolically, people seeing in ultraviolet (a more colourful world) might develop. 3. Conclusion Rapid development and great success is expected in the next 5- 10 years in the field of MEBC. This is seen from the projections given for a number of individual subfields. Contrary to many of the earliest predictions, the present ones have sound bases in the significant progress that is being made in the subfields. Acknowledgments The author is very obliged to the organizers of the Gaithersburg Symposium for the invitation, financial support, excellent arrangements and the kind hospitality, especially to Professor Michael Conrad and Klaus-Peter Zauner. Similar thanks are due to the organizers of the tutorials,

35 (1995) 233-237

231

Professors Harold H. Szu and Felix T. Hong, as well as to the staff of the National Institute of Standards and Technology, in particular Dr Lura Powell. Grant No. 1775(464/91) of the Hungarian National Research Fund is gratefully acknowledged. References Biczo, G., 1994a, Toward room temperature (bio?)molecular computing: I. Background (Tutorial, presented at MEBC 93). Biczo, G., 1994b, Toward room temperature (bio?)molecular computing: II. Perspective of high (room?) temperature realization (Plenary lecture, MEBC 93). Cameron, D.C. and Tong, I-T., 1993, Cellular and metabolic engineering. Appl. Biochem. Biotechnol. 38, 1055140. Conrad, M., 1990, Molecular computing, in: Advances in Computers, Vol. 31, M.C. Yovits (ed.) (Academic Press, New York) pp. 235-324. Hampp, Norbert A., 1994a, Functional optimization of bacteriorhodopsin by genetic engineering and its application in holographic information processing (Plenary lecture, MEBC 93). Hampp, Norbert A., 1994b, Functional optimization of bacteriorhodopsin by genetic engineering and its application in holographic information processing (Tutorial, presented at MEBC 93). Koruga, D., Hameroff, S., Withers, J., Loutfy, R. and Sundaresham, M., 1993, Fullerene C,, (North-Holland, Amsterdam-London-New York-Tokyo). Prigogine, I., 1989, The microscopic meaning of irreversibility. Z. Phys. Chemie. Leipzig 270, 477-490. Stoddart, J.F., 1993, Molecular shuttles on the move. ME News Autumn 15, 4. Traub, J.F. and Wozniakowski, H., 1991, Theory and application of information-based complexity, in: 1990 Lectures in Complex Systems, Vol. 3 of SF1 Studies in the Science of Complexity (Addison-Wesley, Reading, MA) pp. 163- 193.