A few years ago, we made a fairly large bibliometric study in the laboratory - we did not have access to quotation data, but we looked at which of the bioinformatics writes in co -authorship with whom and about what. For various random reasons, its results remained unpublished, but I will tell one of them now. We appreciated all the keywords (Mesh Terms in the Pubmed database) by how their use is changing in each year compared to the previous one. The word is “fashionable” (Vogue), if the frequency of its use grows steadily, or “vintage” (vintage) - this terminology was introduced so as not to offend anyone (after a couple of sentences it will be clear who exactly). Accordingly, the authors can also be classified because they write on fashion or vintage topics.

And it turned out that among the “world experts” (so Evgeny Kunin is recommended on the cover of his book “Logic of the chance”) - bioinformatics with the largest number of citation, with the longest lists of articles and Hirsch indices - he is the only vintage author (for colleagues I will mention that the most following fashion and, possibly, partly forming it - Mark Gerstein and Mark Gerstein and Markiestein and Mark. Per Bork). It seems to me that this is a very important observation. It shows that even in modern vain biology it is not necessary to chase fashion, rushing from epigenetics to metagenomics and from neural networks to protein interactions networks in order to become one of the most influential and respected community members. It also explains why only Kunin could write such a book. I don’t know if he admits to himself, but I’m sure that at heart he pronounced a classic phrase: “But would we swing our Shakespeare William?” Well, that is, on Charles of our Darwin and another half -dozen classics from Fisher and Wright to Mayra and Gulda.

About the content of the book and about the unusual history of its translation into Russian is already described in the reviews of Denis Tulinov and George Lyubarsky, so I will try to talk about what I did not have enough - about the notes of translators and scientific editor. In addition to a pair of little things that should be corrected (see the application to the article below), and the mention of the latest results (partly the author himself does this in the notes to the translation), this would give the possibility of dialogue - as the Biology Direct magazine, one of which is Kunin. In this journal, the author himself makes a decision on the publication, and the article can be published even with negative reviews of the reviewers - but the reviews and answers to them will also be published. The author decides which of the editorial members of the editorial board to offer to write a review, and Kunin, who often publishes his articles in Biology Direct , chooses such reviewers that reading polemics is no less instructive than the article itself. So, Desiderata.

In many places, and even in a special application, Kunin is trying to discuss biological evolution from a physical point of view. At the same time, he completely neglects linguistic analogies. The degree of their depth could be different, but it is strange to ignore the fact that language is another evolving information system, and many problems in its description and study almost literally coincide with problems in the study of the evolution of the genome. Offhand: the boundaries of the language are different languages, and what dialects (cf. determination of the species); the divergence of a single language into a group of related (the origin of Romanesque languages ​​from Latin is a convincing argument in table conversations with creationists, requiring “show an intermediate view between a cat and a dog”); gradual evolution of the language by changing the frequencies of words and other phenomena (cf. the synthetic theory of evolution) and, conversely, relatively rapid restructuring of the language systems, from phonological to syntactic (cf. theory of intermittent equilibrium); hybridization and creolic languages, borrowing (not only words, but also syntactic constructions) and horizontal transfer of genes and opeorons along with regulators; reconstruction of the prisons; coexistence in the language of various codes; contrasting “language and speech” (cf. genome and epigen or, possibly, genotype and phenotype); Finally, the problem of problems is the origin of the language and the origin of life (where some stages can be imagined, but there are enormous holes, for the explanation of which Kunin resorts to the anthropic principle and theory of multiple universes). Of course, there are important differences in both the systems themselves, and in their understanding (say, we, apparently, are better understanding the systematicity of the language than the systematicity of how the genome functions); In linguistics there is the concept of “meaning”, which is difficult to imagine in biology, etc. “But, it seems to me, it would be very instructive to discuss.” It seems that in bioinformatics, as in mathematics, there are two ways to think: physical and linguistic (I will refer to my interviews with Yu.I. Manin and V.A. Uspensky, published in the Triv-Feature, and on the article by Yu.I. Manin, “Languages ​​of mathematics or mathematics of languages”).

The book practically does not have a discussion of the connection of evolution and development-Evo-Devo-and generally quite little is said about the evolution of regulation. Of course, this is due to the author’s own scientific interests and the fact that the successes of bioinformatics in this area are small: the little that we know about the evolution of regulation in eukaryotes mainly come from experimental work. But the swing was not on self-procurement, but on the "third evolutionary synthesis"! One might think that it is the fast evolution of regulatory networks, especially working in the early stages of ontogenesis, leads to sharp changes in morphology, which, in particular, are the basis of traditional taxonomy. In this regard, in the context of discussing the Tree of Life - it would be instructive to discuss what reality the taxonomic levels correspond to. It is clear that not the degree of differences in sequences, but do they exist at all? Formally - if we project a tree of life to the axis of time, will we observe the thickening of the internal nodes? If so, the corresponding branches determine the levels of the family, detachment, class, etc. It seems that in some cases the situation is in this way: say, the difficulties in determining the kinship of mammalian detachments associated with the low lengths of branches at the base of the class prove the reality of both class and detachments. On the other hand, if the branching occurs evenly in time, then all the taxonomy is largely conventionally conventional, which occurs from the arbitrary excretion of some internal nodes as determining taxa. A close topic, examined in detail in the book, but in another context - a comparison of genes. The existence of a large number of genes, specific, say, for chordovs, proves the rationality of their isolation in the taxon. It would be especially instructive to consider the evolution of bacteria from these points of view, which should be close to the author.

Speaking about the models of evolution, it would be interesting to touch the polemic about the existence of group selection, i.e. selection acting at the level of not individual individuals, but groups of related individuals. This theory is designed to explain, in particular, the emergence of altruistic behavior, but is it possible to do without it? A good model is altruistic behavior of unicellular, for which there are several classic examples. Separate cells in the starving colonies of the mixbacteria and mucous membranes slide together and form fruit bodies (see photographs), after which those who find themselves in the “hat” form disputes and scatter in search of a better life, and those who remained in the leg die (by the way, mixbacteria are bacteria, and mucousists - eukaryotes. And a good example of convergent evolution, especially since in both cases the signal molecule is TRAMF). Similarly, in some, in some Bacilli, part of the starving colony performs suicide to serve as a nutrient medium for another part and give them time to go into controversial. In this case, the fate of the cell depends on the concentration of one protein, which is very different in genetically identical individuals for random reasons (cf. Discussion in the book of the role of noise in evolution and the plot about the toxin -antitoxin systems - again, in a slightly different context). In other bacteria, such mechanisms regulate the formation of biofilles, glow, virulence, cellulose degradation, etc. But in unicellular, such behavior is easily explained at the level of individual genes due to the clonal origin of the colonies from one ancestral cell (genetically identical individuals, from the point of view of the selfish gene, is still that one individual, which is selected). To what extent this is transferred to the level of multicellular organisms is a very interesting question.

In conclusion, I must say the main thing. Kunin’s book is a compulsory reading not only for bioinformatics and evolutionists, but, I think, for all biologists. In fact, a research program is declared in it, the depth of which is comparable to classical labors. Even those who are well acquainted with Kunin’s works and already know most of the facts and considerations given in the book will find a lot of instructive in it - at least in how these considerations are collected in a single picture, in the style of writing and the structure of the text. Those who meet this for the first time will find a new way to think about biology, which will undoubtedly affect their own research. The book will also be interesting to non -bioologists, because it shows the advanced land, Frontier of the science of evolution.

1. Evgeny Kunin. Case logic. M.: Center Polygraph, 2014.
2. Denis Tulinov. The evolution of the theory of evolution. TRV-hunger No. 149, 03/11/2014.
3. George Lubarsky. The third evolutionary synthesis. Chemistry and life No. 5, 2014, see also http://ivanov-petrov.livejournal.com/1870801.html .
4. Yuri Manin: "We do not choose mathematics with your profession, but she chooses us." TRV-hunger No. 13, 09/30/2008.
5. V.A. Uspensky: "Mathematics is humanitarian science." TRV-hunger No. 146, 01/28/2014.
6. Yuri Manin. Juits of Mathematics or Mathematics of Languages. TRV-hunger No. 30, 06.06.2009.

Application

As in any review, one cannot do without small amendments and comments. The most significant are collected here.

P. 43: “ Zuckerkandl and Pauling ... proposed the concept of molecular watches: they predicted that the velocity of the evolution of a certain sequence of protein will be unchanged (taking into account possible fluctuations) during long -term temporary intervals in the absence of functional changes .” It seems that the real story is a little more complicated and contradictory. Here is a quote from the article by Emil Tsuckerkandl “The Evolution of Hemoglobin” (collection “Molecules and Cages”, M: Mir, 1966, the original - in the journal Scientific American ): “... In addition to these three postulates, I would like to push the fourth, much more contradictory. I assume that in those modern organisms that differ little from their ancestors, obviously prevailing polypeptide chains, very similar to the polypeptides of their ancestors. Such organisms, a kind of “living fossil”, include a cockroach, a mecch -tail, a shark, and from mammals - lemur. Apparently, many polypeptide molecules synthesized by these organisms are only slightly different from polypeptide chains, synthesized by their ancestors millions of years ago. What is the inconsistency of this postulate? It is often said that evolution lasted equally long both for organisms, which it would seem, differ little from their ancestors, and for those organisms that have changed a lot. Hence, scientists conclude that in their biochemical properties, all these “living fossils” should also differ sharply from their distant ancestors. From my point of view, it is unlikely that morphological signs are preserved in the selection process, but the biochemical properties that underlie them change . ” However, part of the further reasoning of Zuckerkandl, such as assessments of the time of the discrepancy between homologous (now we would say “paralogical”) hemoglobin chains, really relies on the constancy of speeds. But not all: to build phylogenetic trees, he uses the principle, which later became called the “principle of greatest savings”: “ One of the principles of chemical paleogenetics is as follows: when postulating the ancient amino acid residue, assumptions about the lowest number of mutations in the genome, which led to its replacement in the polypeptic chain of descendants ”.

P. 73: " The typical time of the disappearance of the similarities of sequences in homologous genes is comparable with the existence of life on Earth ." It seems to me that there is an ascertainment bias: if some proteins have changed faster, we are simply not able to establish their kinship; This is indicated, in particular, a large number of proteins with the same spatial structure, but sequences similar to random level. On the other hand, for homologists, the discrepancy of which happened very early, we can still observe the differences in the speeds of evolution, and, therefore, their similarity will disappear at different times.

P. 120, about the distribution of degrees of peaks: " Random counts have a bell -shaped distribution of Poisson, and for biological networks the distribution is described by a degree function ." In fact, in several works it was shown that the degree distribution poorly describes biological networks. The fact is that until recently there were no statistical tests to check the hypothesis of the steppe distribution and approval were made by eye-according to the presence of a rectilinear segment in the distribution function built in double logarithmic coordinates (cf. Table. 4-1, right lower graphics). But double logarithmic coordinates - a very insidious thing; Almost any arbitrarily drawn monotonously decreasing function with a monotonous derivative will have such a visually straightforward segment (unless this function is built specifically to refute this statement).

In the discussion of the endosimbiotic origin of the cellular organelles (chapter 7), it may be worth mentioning that, unlike mitochondria, chloroplasts arose at least twice: the primary chloroplast has Paulinella , and it is absent in its closest relatives and, apparently, arose regardless of chloroplast of the ancestors of red and green hydrogen. It seems that the early state of the upcoming acquisition of chloroplast is observed in euclena, which may or not have symbiotic intracellular cyanobacteria: when dividing cyanobacteria, one of the subsidiaries remains, and the second becomes a predator until it acquires a new (before - free -free) cyanobacteria. The question of the border between the organelles and intracellular bacterial endosimbiones of the sucking insects, which can have a very small genome, is even more interesting, comparable to the organelle genome (say, the Carsonella Ruddii genome, Pachypsylla leaflet endosimbionte Venusta , encodes only 182 proteins, and the Tremblaya Princeps genome, one of the Endosimbiones of the powdery and pile of Planococcus Citri, - 121 protein, however, there is another endosimbionte - Moranella Endobia with 406 proteins inside Tremblaya Princeps . I think that the criterion can be exports to the organella of proteins encoded in a nuclear genome.

P. 234: “ The only archeas with more than 5,000 genes were found among mesophils ( namely, some Methanosarcina) , and up to 20 percent of these genomes contain genes of relatively recent bacterial origin .” Indeed, the share of bacterial genes in methanosarcin is greater than in other archi, but the given assessment seems to be high. It was taken from old articles (the beginning of the millennium), and the reason for this error is that at this time the number of sequenced Archei genomes was not enough. Accordingly, when searching for databases, many genes found bacterial, but not archaeal homologs. Reproduction of the procedures used in these works, if it were applied to data banks that change by year, shows that the share of bacterial genes in methanosarcin is monotonously falling (see drawing). A more accurate procedure with the construction of phylogenetic trees for “suspicious” genes leads to an estimate of 6% (Garushyants & Gelfand, Submitted).

When discussing the Lamarkovo model of evolution in Chapter 9, it would be worth mentioning recent work (Dias & Ressler, 2014; Cortiho et al., 2014; Gapp et al., 2014), which shows the inheritance of epigenetic changes and RNA. However, perhaps these works came out after the work on the translation was completed. In general, it would be interesting to discuss the epigenetic mechanisms in the section on genomic complexity in vertebrates in chapter 8, and why only in vertebrates? All mechanisms mentioned on page 271 are also available in insects and plants. А если учесть, что многоклеточность у растений возникла независимо от многоклеточности у животных, можно обсуждать какие-то общие принципы увеличения сложности.

1. Dias BG, Ressler KJ. Parental olfactory experience influences behavior and neural structure in subsequent generations. NAT Neurosci. 2014; 17(1): 89-96.
2. Cortijo S, Wardenaar R, Colomé-Tatché M, Gilly A, Etcheverry M, Labadie K, Caillieux E, Hospital F, Aury JM, Wincker P, Roudier F, Jansen RC, Colot V, Johannes F. Mapping the epigenetic basis of complex traits. Science. 2014; 343(6175): 1145-1148.
3. Gapp K, Jawaid A, Sarkies P, Bohacek J, Pelczar P, Prados J, Farinelli L, Miska E, Mansuy IM. Implication of sperm RNAs in transgenerational inheritance of the effects of early trauma in mice. NAT Neurosci. 2014; 17(5): 667-669 .