|The Logic of Chance: The Nature and Origin of Biological Evolution|
|author||Eugene V. Koonin|
|publisher||FT Press Science|
|summary||An outline of a fundamentally new evolutionary synthesis reflecting key advances in genomics, systems biology and biological physics.|
Secondly, I will preface my review of the technical aspects of this book with my reason for giving it a score of 7 out of 10. The introduction to this book sets very lofty goals. One of them being the hope that this book does for evolution what A Brief History of Time did for physics. That is a seriously tall order and gave me correspondingly high hopes for this book. Koonin, unfortunately, is a very gifted writer and is unafraid of using exceedingly complex sentences such as this gem from page 117 (deliberately taken out of context):
"It has been known for years that a widespread form of global regulation in bacteria is mediated by cAMP, with the participation of diverse adenylate cyclases (a striking case of NOGD); numerous proteins containing cAMP sensors, such as the GAF domain; and the CRP, FNR, and other transcription regulators, also containing cAMP-binding domains."
That sentence is typical of Koonin's writing — lengthy and intricately peppered with many acronyms (only one of which had been described well enough for me earlier in the text). Of course, that paragraph comes with a reference to a paper (like almost all of the paragraphs in this book) from 2010 by Seshasayee so the reader is free to seek external resources if these sentences are daunting.
Considering all of this, I read A Brief History of Time in high school and, despite not having had a physics course yet, learned a lot from it. I attribute that, mostly, to the fact that the sentences are simple and straightforward. Not only that but A Brief History of Time did a great job of building upward from the foundational mechanics of physics while somehow remaining refreshingly brief. This is not the case in The Logic of Chance but I will rush to the book's defense somewhat on that charge. Prior to having read this book, I would have stated my desire that the text start from the basics and work its way up. After reading this book and understanding this field better than I ever have, I now agree that the subject matter of evolution would demand quite the epic tome to accomplish such a feat. I do hope to see future versions of this book with more concise and clear sentences as well as more fundamental concepts explained. If I could have begged Koonin to add one thing to this book, it would be a glossary in the back spanning many hundreds of pages for ignorant readers like myself. Right now this book is for graduate students and academia whereas A Brief History of Time could almost be consumed by anyone who made it through the public school system.
I also sympathize with Koonin's herculean task because modern evolutionary studies seem relatively young compared to other fields like particle physics. As a result, Koonin must (and does) concede in some sections that there still exists largely debated theories. These debates often concern things about which we may never know the absolute truth like the branching factor of a tree of evolution on Earth some indeterminable time ago. As more and more prokaryotes and eukaryotes are added to their statistical algorithms, this may become clearer and yield revelations like the genetic makeup of the last eukaryotic common ancestor (LECA) and free this text of many pages devoted to questions surrounding such origins of life. But for now Koonin must tediously cover all his bases to introduce such things to the reader.
The book starts off by establishing the fundamentals of evolution up until the consolidation of Modern Synthesis. This includes purifying selection, drift, draft, fitness landscapes, etc until Darwinian Evolution was combined with genetics. At this point, the substrate of evolution (the genome) lead to evolutionary genomics. In particular Koonin concentrates on the statistics applied at the molecular level including distance methods, maximum parsimony, maximum likelihood, Bayesian inference and a similar analysis of phylogenetic methods. Koonin establishes early on that evolutionary research can no longer rely merely on phenotypic effects but rather there is a vast array of concrete changes happening at a molecular level.
The book moves on into comparative genomics and discusses extensively the intricate differences between the genomes of viruses, bacteria, archaea and eukaryota. Koonin exhaustively compares these groups through statistics and lays a brief foundation of relationships between genes. From this point on the book is heavily infested with the terminology of homologous, orthologous and paralogous genes. In addition to those the author discusses In-/Out-paralogous, co-orthologous and groups of orthologous (COG) genes. For people unfamiliar with this world, bookmarking and referring to Box 3-1 on page 56 is strongly advised. For the layperson, I believe an expansion of such a graphic would be a great addition to this book. Inside this part, the book also covers a simple but often misunderstood core piece of evolution and that is that evolution has the basic elementary events at the level of gene and genome evolution: substitution, deletion/loss, insertion, recombination/HGT and duplication. Over and over on Slashdot, I see comments that indicate a confusion or perception of evolution being one big monolithic thing. Koonin obviously reads or even studies a lot of other academic fields and tries to explain "the gene universe" as a space-time where there are a few dense clusters of core genes represented in most genomes but most of that space-time is occupied by a huge number of increasingly sparse "nebulae" consisting of rare genes. The author says of this universe: "This organization of the gene universe is distinctly fractal--that is, it appears at all scales of evolutionary distances."
As if that wasn't enough to prove that a definitive phenome narrative (what I alluded to earlier as desired) would be a bad idea, the next section moves on to systems biology and a heavier statistical look at genomics. Beyond the gene status (present or not present) exist two classes of variables: intensive evolutionary variables and extensive phenomic variables. At this point, we're not even talking about tangible things like eye or hair color but rather the underlying mechanisms to those sorts of things like proteins and how they are folded. Everywhere Koonin uses italics, the reader should pay special attention as I found these to be the most interesting key points (example: "Highly expressed genes evolve slowly"). In defining the nature of the evolutionary process, the author covers important concepts like fitness graphs that contain multiple local maxima to demonstrate how non-optimal progressions can occur. Furthermore this section makes it clear that adaptation is not the be-all end-all of evolution. The extensive discussion of the quantifiable properties of genome architecture, functioning and evolution are defined more so by non-adaptive, stochastic processes. Here (and in many later sections) Koonin attempts to use metaphors like Jacob's tinkering and ratchets to help the reader understand these complex concepts but I felt that these metaphors were still so far abstracted that the text could use anything linking these processes to tangible observations in organisms. Again I cannot hold this as a flaw for, after reading the book, it's clear that such a request would be viewed as sophomoric and evidence that I am unable to progress past The Origin of the Species (this book's key objective).
Koonin then moves on to the prokaryotic world and examines their genes and operons while paying special attention to an odd case: cyanobacteria. Most importantly in the prokaryotic domain, extensive comparative genomics has revealed a concept called horizontal gene transfer (HGT). I was personally hoping that Koonin would seize upon this novel concept and its importance in bacterial antibiotic resistance and how bacteria can evolve to dissolve novel compounds. For better or for worse, Koonin sticks to the pure purpose of this book and extensively covers important HGT discoveries like the convergence of protein sequences in similar groups of bacteria and archaea. Some selfish genes rely so heavily on horizontal mobility that they are dubbed "mobilomes" and Koonin discusses their aspects extensively. Darwin's Tree of Life concept was a very small eukaryotic part of the big picture that Koonin tries to re-invent as the "Forest of Life" or "Web of Life" (considering HGT). A whole chapter is devoted to discussing its properties and graphically visualizing its structure based on extensive surveys and what we know today.
From there the author discusses the origins of eukaryotes, Last Eukaryotic Common Ancestor (LECA), the branching factor of its evolution, its relative distance to the point of symbiogenesis in proposed evolutionary trees and the many competing theories about that tree. This section of the book spends considerable time examining the inferred origins of basic eukaryotic cell functioning and also discusses at length the archaeal roots of elaborate systems with the exception of the mitochondrion. This chapter also looks at the perplexing features of introns in eukaryotic genes. Koonin then tackles the misconceptions and abuses of the word complexity in all aspects of evolution. He applies information theory to the genetic code and notes that "information (entropy) tells us very little about the meaningful information content or complexity of a genomic sequence." It is then suggested that a new way to compute entropy and complexity is to examine the alignment of orthologous sequences instead of single sequences. For people interested in information theory, chapter eight is the most fruitful where Koonin proposes a computable formula for biological (evolutionary) information density. Like Claude Shannon's ability to infer many important aspects of communication, Koonin's modifications allow us to calculate that perceptually complex organisms possess more "entropic" genomes while perceptually less complex organisms like bacteria have the tightly packed and information dense "informational" genomes. After establishing these studies in information theory, Koonin is able to argue that neutrality of mutations that are fixed during evolution is the null hypothesis for all molecular evolutionary theories. All of this aids the author in discussing why evolution progressed passed single celled organisms that already had 1,000 to 1,500 genes to larger sets of genes in multicelled organisms.
Chapter nine tackles the modalities of Darwinian, Larmarckian and Wrightean evolutionary theories. This chapter improves upon the simplistic triad of heredity-variance-selection that defines Modern Synthesis by showing that the relationship between population size and environmental stress determines which of the three modalities is expressed the most in evolution while at the same time observing the importance of entropy (noise) at all levels of transmission. Koonin shows that by combining very well known molecular mechanisms we can achieve a complex scenario like Jean-Bapteste Lamarck's proposed modality of evolution. The text gives viruses the same treatment which, despite my assumption that they would be easier to analyze, appear to have many of the same complexities that prokaryotes and eukaryotes have. Possibly even more so given the effects of the Red Queen Hypothesis and all of the counterdefense genomes in some viruses. Furthermore the cellular empire and virus empires have two-way exchanges of genes. The truth is we know very little about the virus world — considering its size and history — and the author postulates that viromes in unknown and unstudied viruses consist largely of uncharacterized "dark matter" (again, borrowing terms from cosmologists).
Koonin then approaches the next logical step backwards: the last universal common ancestor (LUCA). He starts by listing the arguments that cellular life indeed had a common ancestor and looks at competing theories (for example cell organization complexity versus genetic complexity leading to different models of varying degrees of cellularity). In chapter twelve, Koonin covers the topic that is often the hardest to imagine — the origin of life. This is interesting and particularly difficult because the translation system itself at some point evolved. Interestingly enough, these 60 protein-coding genes and ~40 structural RNA genes are the only complex ensemble of genes that are conserved across all extant cellular life forms. So, of course, the point in the evolutionary tree where this had developed is discussed as well as the Darwin-Eigen cycle. The latter requiring a system of a far greater complexity in order to be started. So the author begins examining the proposition that over time and due to their catalytic properties ribozymes lead to processive synthesis of peptides (long enough to be the first proteins). After discussing the eleven stages this would have to encompass, the author discusses the existing skepticism of models that try to explain how replication and transcription came about. This chapter also tackles geochemical and chemical propositions on the origin of life — something that has been discussed on Slashdot before. This research centers on networks of inorganic compartments consisting of catalytic surfaces with gradients of heat and acidity that could have supported primordial organic chemistry.
The book ends with a chapter devoted to reiterating topics as well as asking important questions like whether or not another biological evolution model is necessary/feasible as well as caution against logic like the progress fallacy or criticizing a concept like "the selfish gene" because it sounds "undignified." Though these are tempting arguments because of their simplicity, they have proven fruitless. A diagram on page 412 reminds us just how complex the flow of genetic material is between the virus empire and the cellular empire.
There are two appendices to this book and, perhaps because they use a softer language, they were much more accessible to me yet posed more questions than answers. Appendix A concentrates on the philosophy of postmodernism, the infeasibility of synthesis and the distrust of metanarratives. The author argues that any paradigm presented must include oversimplification and that we merely replace them with better metanarratives. It is also important to ask these questions about the current paradigms for without them we would never have come up with drift, draft and various neutral ratchets to improve old models. Koonin references Hawking and Mlodinow with the concept of model-dependent realism which stresses that scientists merely construct models that are in turn swapped out for better models given how well they explain data and predict the outcomes of experiments. Lastly Koonin refers to Popper's famous falsification paradigm and his subsequent position on how invaluable evolution is purely on the grounds that it arms us to model and understand specific experiments. The second appendix deals with roughly estimating the probability of life arising given inflationary cosmology. I know this back of the envelope math has become popular given recent discoveries of exoplanets in the news but I felt the few references to the "many worlds in one" model deserved to be placed in a separate book. Nevertheless, Koonin covers both the strong and weak forms of the anthropic principle and looks at the connotations they hold for evolution.
The references at the end of this book are extensive — 38 pages of two line references. It should probably be mentioned that Koonin's references to his own work consist of two of these pages although at no point did it sound like he was unfairly proffering his theories over others. At certain points I had to wonder whether or not I was reading a lightly adjusted abstract from a peer reviewed paper or a book. This is most evident in one of the figures of an appendix on page 437 that reads "This is a formulation of the 'weak' anthropic principle adopted for the context of this paper." Since it is a graphic and in the appendix, it's forgivable but caused me to wonder if the rest of the book couldn't be more seamlessly tied together with transitionary language for novices like myself. Amazingly, I found maybe one grammatical error and no typos in this book which was a refreshing experience for a first edition. Also, this is one of the best bound books I've had the pleasure of reading, its spine has held up to hours of laying it flat open while I googled for a better understanding. While $50 is pricey, the book is built to last and this $10 premium over the kindle edition is worth it if you must hold a physical copy of a book. It saddened me to be reminded that some states struggle with including the core concepts of Darwinian evolution anywhere in their K-12 curriculum. And should those students desire to break new ground in this modern field, texts like The Logic of Chance are that much further away from them.
You can purchase The Logic of Chance: The Nature and Origin of Biological Evolution from amazon.com. Slashdot welcomes readers' book reviews — to see your own review here, read the book review guidelines, then visit the submission page