V. V. Anshelevich

Computer Simulation of DNA Supercoiling

We treat supercoiled DNA within a wormlike model with excluded volume. A modified Monte Carlo approach has been used, which allowed computer statistical-mechanical simulations of moderately and highly supercoiled DNA molecules. Even highly supercoiled molecules do not have a regular shape, though with an increase in writhing the chains look more and more like branched interwound helixes. The averaged writhing (Wr) approximately 0.7 delta Lk. The superhelical free energy F is calculated as a function of the linking number. Lk. The calculations have shown that the generally accepted quadratic dependence of F on Lk is valid for a variety of conditions, though it is by no means universal. Significant deviations from the quadratic dependence are expected at high superhelical density under ionic conditions where the effective diameter of DNA is small. The results are compared with the available experimental data.

Torsional and Bending Rigidity of the Double Helix from Data on Small DNA Rings

We have calculated the variance of equilibrium distribution of a circular wormlike polymer chain over the writhing number, [Wr)2), as a function of the number of Kuhn statistical segments, n. For large n these data splice well with our earlier results obtained for a circular freely jointed polymer chain. Assuming that [delta Lk)2) = [delta Tw)2) we have compared our results with experimental data on the chain length dependence of the [delta Lk)2) value recently obtained by Horowitz and Wang for small DNA rings. This comparison has shown an excellent agreement between theory and experiment and yielded a reliable estimate of the torsional and bending rigidity parameters. Namely, the torsional rigidity constant is C = 3.0.10(-19) erg cm, and the bending rigidity as expressed in terms of the DNA persistence length is a = 500 A. The obtained value of C agrees well with earlier estimates by Shore and Baldwin as well as by Horowitz and Wang whereas the a value is in accord with the data of Hagerman. We have found the data of Shore and Baldwin on the chain length dependence of the [delta Lk)2) value to be entirely inconsistent with our theorectical results.

Effect of Excluded Volume on Topological Properties of Circular DNA

We have performed computer simulations of closed polymer chains with allowance for the excluded volume effects within the framework of the free-joint model. The probability of knot formation, the linking probability of a pair of chains and the variance in the writhing number proved to be significantly affected by the excluded volume effects. This is true even for DNA with completely screened charges for which the b/d ratio (where b is the Kuhn statistical length and d is the diameter of the double helix) is as large as 50. Allowance for the electrostatic repulsion (change of the DNA effective diameter d) further increases the effects. The most dramatic dependence on d is found for the probability of knot formation. The data on the dependence of the variance of writhing, mean value of (WR)2, on d indicate that the DNA superhelix energy should be significantly ionic strength-dependent. Special calculations have shown that the free-joint model underestimates the mean value of (Wr)2 value by about 20% as compared with the wormlike model.

Symmetric random walks in random environments

We consider a random walk on thed-dimensional lattice ℤd where the transition probabilitiesp(x,y) are symmetric,p(x,y)=p(y,x), different from zero only ify−x belongs to a finite symmetric set including the origin and are random. We prove the convergence of the finite-dimensional probability distributions of normalized random paths to the finite-dimensional probability distributions of a Wiener process and find our an explicit expression for the diffusion matrix.

A hierarchical approach to computer Hex

Hex is a beautiful game with simple rules and a strategic complexity comparable to that of Chess and Go. The massive game-tree search techniques developed mostly for Chess and successfully used for Checkers and a number of other games, become less useful for games with large branching factors like Hex and Go. In this paper, we describe deduction rules, which are used to calculate values of complex Hex positions recursively starting from the simplest ones. We explain how this approach is implemented in HEXY – the strongest Hex-playing computer program, the Gold medallist of the 5th Computer Olympiad in London, August 2000.

Neural Network Models for Promoter Recognition

The problem of recognition of promoter sites in the DNA sequence has been treated with models of learning neural networks. The maximum network capacity admissible for this problem has been estimated on the basis of the total of experimental data available on the determined promoter sequences. The model of a block neural network has been constructed to satisfy this estimate and rules have been elaborated for its learning and testing. The learning process involves a small (of the order of 10%) part of the total set of promoter sequences. During this procedure the neural network develops a system of distinctive features (key words) to be used as a reference in identifying promoters against the background of random sequences. The learning quality is then tested with the whole set. The efficiency of promoter recognition has been found to amount to 94 to 99%. The probability of an arbitrary sequence being identified as a promoter is 2 to 6%.

On the ability of neural networks to perform generalization by induction

The ability of neural networks to perform generalization by induction is the ability to learn an algorithm without the benefit of complete information about it. We consider the properties of networks and algorithms that determine the efficiency of generalization. These properties are described in quantitative terms. The most effective generalization is shown to be achieved by networks with the least admissible capacity. General conclusions are illustrated by computer simulations for a three-layered neural network. We draw a quantitative comparison between the general equations and specific results reported here and elsewhere.

Application of Polyelectrolyte Theory to the Study of the B-Z Transition in DNA (1)

We have used the polyelectrolyte theory to study the ionic strength dependence of the B-Z equilibrium in DNA. A DNA molecule is molded as an infinitely long continuously charged cylinder of radius a with reduced linear charge density q. The parameters a and q for the B and Z forms were taken from X-ray data: aB = 1nm, qB = 4.2, aZ = 0.9 nm and qZ = 3.9. A simple theory shows that at low ionic strengths (when Debye screening length rD much greater than a) the electrostatic free energy difference FelBZ = FelZ - FelB increases with increasing ionic strength since qB greater than qZ. At high ionic strengths (when rD much less than a) the FelBZ would go on growing with increasing ionic strength if the inequality qB/aB greater than qZ/aZ were valid. In the converse case when qZ/qB greater than aZ/aB the FelBZ value decreases with increasing salt concentration at high ionic strength. Since X-ray data correspond to the latter case, theory predicts that the FelBZ value reaches a maximum at an intermediate ionic strength of about 0.1 M (where rD approximately a). We also performed rigorous calculations based on the Poisson-Boltzmann equation. These calculations have confirmed the above criterion of nonmonotonous behaviour of the FelBZ value as a function of ionic strength. Different theoretical predictions for the B-Z transition in linear and superhelical molecules are discussed. Theory predicts specifically that at a very low ionic strength the Z form may prove to be more stable than the B form.(ABSTRACT TRUNCATED AT 250 WORDS)

Laplace operator and random walk on one-dimensional nonhomogeneous lattice

A classical result of probability theory states that under suitable space and time renormalization, a random walk converges to Brownian motion. We prove an analogous result in the case of nonhomogeneous random walk on onedimensional lattice. Under suitable conditions on the nonhomogeneous medium, we prove convergence to Brownian motion and explicitly compute the diffusion coefficient. The proofs are based on the study of the spectrum of random matrices of increasing dimension.

A Theoretical Study of Formation of DNA Noncanonical Structures Under Negative Superhelical Stress

The development of statistical mechanical models of the formation of noncanonical structures in circular DNA and the finding of the energy parameters for these models made it possible to predict the appearance of such structures in a DNA with any given sequence. It does not seem feasible, however, to perform such calculations for DNA sequences of considerable length by allowing for all the possible states. We propose a special algorithm for calculating the thermodynamic characteristics of various conformational rearrangements in DNA that occur under negative supercoilings, allowing for several possible states of each base pair in the chain. Calculations have been performed for a number of natural DNAs. According to these calculations, the most likely noncanonical structures in DNA under normal conditions are cruciform structures and the Z form. The results of the calculations are compared with the experimental data reported in the literature. State diagrams have been computed for a number of inserts in circular DNA that can adopt both the cruciform conformation and the left-handed helical Z form.