This is a book about conversion and expression. It is a clearly written textbook that teaches the reader how to convert biomolecular observations into mathematical models of genetic circuits. The author also shows how to use such models to discover the principles of circuit operation with suitable resolution to predict, control, and design function. These circuits, encoded in the one-dimensional information of the genome sequence, are expressed by the cellular machinery into the teeming variety of three-dimensional life with which we are familiar. Aside from evolution itself, I can think of no more awe-inspiring transformation.
Evolution implicitly plays its part in this book. The fact that there are generalizations and abstractions to be made about the architecture and organization of the cellular machinery is traceable to the mechanisms of evolution. From the origin of life approximately 3.5 billion years ago, via the extraordinary mechanisms of mutation and genetic transfer, the genome has explored a vast space of possible configurations of cellular networks that have been subject to natural selection. “Designs” that lead to efficient procreation survive with the rest eventually weeded out. There is a good argument that natural selection and the mechanisms of mutation incidentally leads to evolvability—the ability of a design to withstand and be quantitatively tuned by small variation and, in addition, have significant numbers of routes to reconfiguration for new, plausibly useful, function. Evolvability itself may lead to the proliferation and reuse of “motifs” of function ranging from macromolecular signal sequences and domains to whole subnetwork architectures. That evolution may drive towards recurrent molecular “designs” and circuit motifs is one of the key observations underlying the promise and power of Systems Biology. To those who, by profession, appreciate and attempt to make sense of the function of complex systems, life and the biological circuits that underlie it provide a nearly irresistible allure. It is increasingly clear that there are principles of organization, architecture, and (evolutionary) design in biological circuits, but there are many details and deep principles yet to be discovered.
This book is also about the conversion and expression of the author, who is a world-class engineer and has also written an excellent textbook on Asynchronous Circuit Design for electronics. I am not an electrical engineer, but as I understand it, engineering with a clock, enforcing synchronous operation of a circuit, greatly simplifies the design of circuits since it is predictable when (and where) signals are coming from. Thus, the timing of arrivals of signals is, at most, a
second order consideration. However, as such circuits increase in complexity and size, it becomes impractical to ensure that even the signal from the clock itself is received simultaneously by all elements. The assumptions of synchrony breakdown leading to immense complication in design. Professor Myers is an expert in controlling this complication. Biological circuits do not really have a central clock. Worse, they do not adhere to homogeneous Boolean abstractions nor are all the inputs, outputs, and internal mechanisms unambiguously known even for relatively small subsystems of cellular function. One can see how an engineer of Professor Myers’ pedigree might be drawn to this different and difficult type of circuitry.
Professor Myers’ book is really about Systems Biology. While the origins of the term are a bit cloudy, it is fairly clear that it arose as an identifiable discipline in the mid-sixties though the intellectual foundations I would argue came quite a bit earlier. But the 1960’s is when both the genetic code was cracked and the molecular basis for the observation that the expression of genes and the activities of their encoded proteins could be regulated by their own and other genes’ products was discovered. This “Central Dogma” was solidified—DNA is transcribed into RNA and RNA is translated into protein, and it was clear these macromolecules formed complex dynamic networks of interaction. It became equally clear that the study of how these lines of connection logically translated nearly static genotype to dynamic phenotype was a large and sophisticated study in itself. This kind of approach is in contrast to the powerful and prerequisite discovery science that is still the dominant paradigm in biology. Once the parts of the system are known and the mechanisms of their individual interactions are characterized, there is still the problem of understanding how that system, so exquisitely described, functions.
Imagine that I’m handing you the schematic for one of Professor Myers’ asynchronous digital circuits complete with a description of how every component worked. How would you determine what it does and how it encodes and transforms its signals into dynamical calculations? What are the key parameters that control this function, and whose variation might lead to failure or change of behavior? How would you understand the principles of its function? (After all, there are many circuits that might implement a given function—how does this one work and why was this design selected?) How would you uncover the structure of its design? Is the system “modular” in expression or control? Are there subcircuits that themselves have useful function? Are there elements and architectures that confer robustness to uncertainty in the environment? How can we perturb the circuit to move it from a possibly unknown state to a desired one? For biological circuits, there are
also questions that skate dangerously close to teleology (I am a not-so-closeted teleologist myself). Is the circuit optimal for something identifiably “engineering”? By its existence, a biological system must be very good at surviving and procreating, but how? Is energy usage or temporal response optimal in some way? Is the particular architecture necessary and/or sufficient for a particular function such as, for example, gradient sensing? Does it implement a winning strategy in a discoverable evolutionary game? Is the architecture especially amenable to “upgrade” or use of pre-existing parts from other devices? Or is it an accident of largely vertical inheritance from progenitors that arose under different conditions?
I believe these questions, applied to biological circuits, are the particular but non-exclusive bailiwick of Systems Biology. The specific experimental tools, theoretical frameworks, and computational algorithms that apply to these problems are changeable and distinct. The choice of tools, of course, depends on the resolution or scale of the question under consideration—and technology in this area is changing incredibly quickly. It is not the tools but the questions
about the integrated function of cellular networks that define Systems Biology in my mind. If molecular or mechanistic discovery comes along for a ride, that is a welcome bonus. That the focus is on circuit behavior and (evolutionary) design is what so draws engineers like Professor Myers to the field.
When Chris and I met at Stanford in the mid-nineties, I was working to understand how the physical mechanisms of the cell might have evolved to implement comprehensible “engineering” function, and how such function might be inferred from data. Chris was an engineering student working on systems complex enough to require breaking the standard electronic engineering paradigm and developing new principles of circuit design. With our mutual friend, Michael Samoilov, we began an exchange that continues today, learning from each other and sharing our appreciation for the designs of Dawkin’s old Blind Watchmaker, evolution.
I will claim credit for introducing Professor Myers to the venerable and very fascinating Phage Lambda as the model system of choice to test nearly all ideas about Systems Biology. Phage Lambda was involved in an astonishing number of fundamental discoveries in genetics, biochemistry, virology, and development. Lambda has stood, since the inception of its study mid-last century, as one of the few biological systems that has strongly linked physical, engineering, and biological
scientists. The impact of Lambda and its mechanisms on biology and biotechnology would be hard to overstate. At its heart are two famous decisions: the decision upon infection to either create many progeny, thereby killing the host (lysis), or to integrate into the host chromosome (lysogeny), thereby conferring immunity to further infection; and the decision, after integration, to re-enter the lytic path if host conditions become unfavorable. Birth, death, predation, temporary symbiosis, and decision-making are all encoded in one 50 kb genome package (ignoring host functions). Together with my colleagues, Harley McAdams and John Ross, I was deeply immersed in models of its networks to study the implications of molecular discreteness and noise in the robustness of cellular decisions. For an engineer like Chris, the famous “switch” was a clear entry place and provided a common point of reference for our discussions. Chris took off on his own, deepening his understanding of Systems Biology through research on this model system. It is gratifying to see Lambda as an organizing example used so effectively in this text.
One of the defining properties of the study of Systems Biology—even for this relatively simple system—is that no one can agree on what one needs to know to study it effectively. Even more than half a century after its discovery, there are still new mechanisms being discovered about the functioning of the Lambda switch. Thus, the classical biologists demand (rightly so) that anyone studying the switch must be an expert in biology and understand Lambda lore, in particular, quite well because otherwise any models of its architecture and function will not properly capture the mechanistic complexity or uncertainty about its operation. The biochemists and biophysicists will insist that the particulars of the mechanisms—how exactly the proteins interact and degrade, how DNA loops to form a key component of the switch, how the complex thermodynamics of multicomponent promoter binding leads to the proper ordering of states for stabilizing the two decisions, how the stochastic effects due to the discrete nature of the chemistry leads to diversity in behavior—all must be considered carefully or formally discounted to justify explanations for how the system works. The computational biologists and systems biologists require, on top of the above, deep understanding of biological data analysis, numerical algorithms, graph theory, and dynamical systems and control theory in order to build appropriate models justified by data and to analyze
them properly, and a fine sense of how to interpret and abstract the principles of control buried in the dunes of detail about the system.
Students, throw up your hands! How do you start with all this deep and complex science and engineering to learn? As a first step, we must learn to live with our
ignorance, and only then can we begin to fight to eradicate it. The frontier of science is where ignorance begins, and for Systems Biology, there is plenty of
undiscovered land. This book by Professor Myers is one of the few texts in the area that gently brings the uninitiated to these edges. I congratulate him for his achievement—Engineering Genetic Circuits admirably touches on much of the “required” knowledge above while creating a minimal toolset with which beginning students can confidently venture into this exciting new territory of Systems Biology.