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 ionic strength dependence of the cooperativity factor for DNA melting.

The melting temperature for the d(AT)24.d(AT)24 stretch, located inside the DNA helix and terminally, have been determined in a wide range of ionic strength values (0.01 - 1 M Na+).

Allowance for heterogeneous stacking in the DNA helix-coil transition theory.

Melting profiles of eight DNA molecules with lengths ranging from 849 to 4362 bp have been measured in an SSC buffer where the melting is an equilibrium process up to complete strand separation. A theoretical analysis shows that the melting profiles depend on only eight invariants that are linear combinations of 10 original stacking parameters. As a result it is impossible to determine the 10 parameters from the melting profiles.

Determination of DNA cooperativity factor.

The paper presents measurements of the difference in the melting temperature of a colE1 DNA region when it is located inside the DNA helix and at its end. A direct comparison of calculations based on the rigorous theory of helix-coil transition with experimental data for .2 M Na+ (the conditions for fully reversible melting) yielded the value of 2.