| The kinetics and thermodynamics of loop formation by segments of double-stranded DNA was studied by computer simulation. First, we studied a general problem of polymer physics, we estimated the effect of the impenetrability of the chain backbone on the dynamic properties of very long chain. Three models of polymer chains were compared: phantom ones with and without excluded volume and a non-phantom model with excluded volume. The end-to-end distance relaxation time, taum, and the time of diffusion-controlled cyclization, taua, were computed directly using Brownian dynamics. In accordance with known theoretical results, we observed that for long chains taua is directly proportional to taum. We determined that the chain impenetrability itself has small effect on the magnitudes of taum and, consequently, taua. They exceeded the values for phantom chains by 50%. Second, we studied kinetics and thermodynamics of loop formation by very short DNA using Brownian dynamics. DNA was modeled as a discrete wormlike chain. A new Monte Carlo algorithm was developed for efficient calculation of the probability of loop formation and cyclization for short DNA. Our results showed that the formation of small loops is a very slow process: for DNA fragment shorter than 50 nm taua can be comparable to the lifetime of the living cell. Third, we developed a rigorous quantitative approach to the analysis of ligation of short DNA oligonucleotides with sticky ends into linear and circular multimers of various lengths. Such experiments are used to determine structures of the oligonucleotide containing specific chemical modification, bound ligands, or irregular structural elements. By numerical simulations of the reaction, we found a procedure to determine j-factors for different multimers from the distribution of the reaction products. We, then, found empirical equations that connect the extracted j-factors with the structural parameters of the oligonucleotide, such as its bend angle and rigidity. To obtain the equations, we used the new Monte Carlo procedure and computed j-factors for a large array of conformational parameters. |
