Now, it turns out that my connection with genetics is, in fact, genetic. What I mean here is that it has to do with my father, the late Professor Herman Mark. Many of you here today met him during the years that he lived with us at the Bauer House in Austin so you know who I am talking about. Anyway, let me explain.
II. Family Genetics
In 1925 while working at the Institute for Textile Studies at the University of Berlin, my father and his colleague, J. R. Katz, were the first to use x-ray diffraction to establish the techniques to determine the structures of large molecules of biological origin. People at the Institute were interested in establishing the physical properties of fibers used by the German textile industry. Mark and Katz pioneered the use of the then new physical technique of x-ray diffraction to initiate the determination of the molecular structure of cellulose (Ref.1). They chose cellulose because that compound is the major constituent of cotton, which is the most widely used and versatile textile fiber. Later, when the structure was finally determined by Mark and Meyer, it was found that with this knowledge predictions could be made about the behavior of the fibers and the textiles made from them. It was a useful technique that substantially improved the textiles manufactured in Germany and it led to much wider applications of x-ray diffraction (Ref.2). Hopefully, this begins to establish my genetic credibility but I still have to explain the bit about Nobel laureates.
During his academic career, my father supervised two Ph.D. students who later were awarded Nobel prizes. One was Eugene Wigner, whose Ph.D. work had to do with the use of x-ray diffraction to establish the crystal structure of rhombic sulfur (Ref.3). He received his Nobel Prize for the application of group theory to the solution of problems in quantum mechanics, a theory he learned while analyzing the x-ray diffraction patterns produced by the Sulfur crystals he examined to earn his Ph.D. The other student was Max Perutz, and he is the one who is important for our discussion today because he was the one who developed the techniques that were later used by James Watson, Francis Crick, Maurice Wilkins, and Rosalind Franklin to determine the structure of deoxyribonucleic acid (DNA). In that sense, Max might be called the grandfather of the genetic revolution! In any event, I knew both of these gentlemen quite well because of my father, so I have at least some experience in talking to Nobel laureates.
Max Perutz finished his Ph.D. thesis in 1937 doing research on a problem in classic organic chemistry at the University of Vienna. Unfortunately, the thesis was never published because both Perutz and my father left Vienna shortly after the work was completed. My father saw the danger posed by the Nazis at that time since he was himself planning to leave Austria. Thus, he urged his young protÈgÈ to leave Austria as well, preferably sooner rather than later. There was a vacancy in the laboratory of Sir William Lawrence Bragg at Cambridge University. At the time, Bragg was the world's leading expert on x-ray diffraction. Perutz was worried that he knew nothing about x-rays whereupon my father said: "You will learn," and learn he did! After a painstaking series of experiments using much trial and error, Perutz succeeded in determining the molecular structure of myoglobin and later of hemoglobin, a feat for which he was awarded the Nobel Prize in 1962 (Ref. 4). Hemoglobin was a much more difficult proposition than cellulose. It is a very complex three dimensional structure and Perutz had to go far beyond what was done with cellulose, which is a relatively simple structure that has cylindrical symmetry. Perutz completed this work a few years before the structure of DNA was established using the same techniques in 1954 (Ref. 5).
III. Computers and Their Limits
The realization that DNA was the template that determines the properties of all living things using a three-digit code was the intellectual breakthrough that brings us here today. You will hear both about the promise and the problems raised by the new insights that have been gained. The prospects are truly as mind boggling as they are varied. Instead of making lists, let me talk about what interested me most about the genetic revolution. Not surprisingly, it is related to engineering. In 1979, when I was serving as Secretary of the Air Force, we began to wonder how much computer capacity we could put on the Airborne Warning and Control System (AWACS) aircraft. I called my old friend Walt Morrow, who was director of the MIT Lincoln Lab at the time, and asked him what he thought about the question. He replied that it has to do with the size of switching elements. Transistors that control the flow of electrons cannot be smaller than a certain size. They depend for their operation on the establishment of energy levels called Brillouin zones and, if the transistor becomes too small, then these zones can no longer exist. By doing some elementary quantum mechanics, Walt estimated that the smallest transistors must have at least 1012 atoms to work--that is one thousand billion. We are today about a factor of 50 away from this limit, the smallest current switches today contain about 5 x 1013 atoms. From a practical viewpoint, there is still lots of room for improvement. Every factor of two is worth lots of money so don't sell your computer company stocks! However, from a scientific viewpoint, Walt could foresee even twenty years ago that at some point, we would come up against a fundamental limit.
IV. Biological Switches
The final question that I asked Walt was what he thought could be done to overcome that limit and he then told me a most interesting story. He said that there were people in the biological departments at MIT who were beginning to learn how the switching elements in our nervous system worked. He told me that these switches were molecules containing about ten thousand atoms that are impregnated on the membranes enclosing nerve cells and that these molecules control the flow of ions (and hence electric currents) in our nervous system. If we could learn how to use and control these biological switches, Walt told me, we could build computers that are smaller by many orders of magnitude than the ones we have today. To be more precise, biological switches are eight orders of magnitude smaller than electronic switches. They are much slower since they operate on currents of relatively slowly moving ions rather than on rapidly moving electrons. However, we have since learned that it is relatively easy to overcome this problem by introducing what we call parallel computer architecture (Ref. 6). We have made much progress in learning about biological switches and what my computer friends call neural networks since I had my conversation with Walt Morrow almost twenty years ago. Walt will very probably turn out to be right that we will overcome the quantum mechanical limit imposed by size on transistors by the use of biological switches.
V. A Speculation
It is this last point that leads me to the speculation with which I want to wind up this introduction. The human mind is by far the most complex and sophisticated computer we know about. It has about 1011 (one hundred billion) synapses and each of these may contain hundreds of thousands of switching molecules, so astronomical numbers of switches exist in the brain-- approaching say 1016. Compare this to our best current computers which have about a hundred million (109) switches. The point of all this is that the way in which the switches in the human mind are wired up is contained in the genetic code. The architecture of the computer that is the human mind is hidden somewhere in the three plus billion units of the DNA molecule. By knowing the complete sequence of human DNA, will it become possible to build computers that approach the capability of the human mind? Can we understand how the architecture of the brain is encoded in the DNA molecule? If so, can we replicate the principles and use them to build artificial brains? Can we put a bunch of nerve cells in a Petri dish and then, by somehow introducing the information contained in the DNA, wire them up in such a way that they become a computer more capable than anything we now possess? My own judgement is that all of these things will eventually become possible. I believe that this is the prospect before us and the things we will discuss today only represent the first step in the Genetic Revolution.
1. J.R. Katz and H. Mark, Z. Electrochem, vol. 31, p. 105 (1925).
2. K. H. Meyer and H. Mark, Ber. Deut. Chem. Ges., vol. 61B, p. 593 (1928).
3. H. Mark and E. Wigner, Z. Phys. Chem., vol. 111, p. 398 (1924).
4. M.F. Perutz, Proc. Roy. Soc. A, 195, p. 474 (1949).
5. J.D. Watson and F.H.C. Crick, Nature, 171, p. 737 (1953).
6. R. Michael Hord, The Illiac IV, The First Supercomputer, Computer Science Press, Rockville, MD (1982).